CA2000700A1 - Signal processing method and apparatus for flowmeters - Google Patents

Abstract

Disclosed is a method of measuring the flow of a fluid in a pipeline, wherein the pipeline has noise, comprising the steps of: a) inserting a flow barrier in the pipeline, thereby forming a wake of vortices; b) creating an electronic signal for each vortex by means of a device which also produces electronic signals for the noise; c) transmitting the electronic signals of the vortices and noise to an electronic processor; d) converting the electronic signals of the vortices and noise vibrations from the time domain to the frequency domain; e) removing the electronic noise signals from the electronic vortex signals by means of cross-correlational computations, thereby computing the vortex frequency; and f) computing the flow from the computed vortex frequency.

Description

C T ~ ~ ~ T H U 1 ~ ~ æ Z~0~700 P . ~3 5 BACi;GROUND O~ THE I~VENTlO~
I . pield Or the I nvention The present inventjon rel~te~ to rlclwmeter~ and digital signal pr~essin~.

2 t~tion of the Prior Art l h~ v~rleY ~hedding flo~meler is emerging as onc of th~ mo~l populRr rlo~me~ers. No moll~ing par~s are subjected to wear ~nd te~r. Il ha~ a rlJgged and sturdy construction suitabie ror applications involving ex~reme temperatures and pressures. Il pro~ide~ a highly accurate and reliable fk~
measurement on the voll~me flo~ rate o~ al~ ~ype~ of fluids. In~eed, the vorte~ shedding rlowmeter has the potential o~ the best ~ind of flo~met~r rOr n any applications if several persistent problems can be ~olve~ me Or ~hich are ~ol~ted by this invention.
~ 'orte~ sheddin8 is 8 nalural phenon~enon tha~ occurs when a rl~lid pn~se~ ~ non-streamline~ body. Such bodie~ (blurr bodies) include smc-l;es~acks, flsgpoles and bodies in~erted intc~ rlowing fluids in pipeline~
Vortices ~re locali~e~ zc~nes of inqeased velocity. They ~orm in ~he fl-lid stre~m at Ihe ini~ialing or bluff b~dy. ~hese vortices detach them~elves from the initiating body ~nd are "shed' ~ownstream in a "slreet .
20 l`he ~t~rlice8 allernately shed from the oppo~i~e sides Or the initiating body at a rale ~ hich is direclly proportional and linear ~ith the fl~ fl~w When the non-stre~mlined bbjed or blulrr body is placed in the fluid path inside a pipeline, 1his same principle can be ulsed to detect the flo~rale Or the fluid. By counting ihe vortices shed f!om the blurr body over a period Or lime, one can compute the flo~rate, and, by u~e Or the pipe dimen~ n~

O C T -- 1 ? -- 8.9 T H 1 1 1 ~ ~ ~ z~JQ70o p ~ ~

~ J 1, and flui~ ch~l acteri~llcs~ une can ~Iso eompu~ Ih~ v~ elric rluid fln~ in the ~ir~ itselJ. Sll~.h teGhniqu~ an~ ~.omputatic~n~ are well-~;no~n ~0 per~ons ~killed in this fi~ld In some vorlex flo~meter structyres. ~uch as that de~crited in Li~.
~alenl llc). 4,~9~,0}2 isslJéd to H. Lew, et 8i., a wing is pl~oed downstream ~rom the bluff body. 'rhe wir~g mo~e~ or Yibrales fr~m side-t~-side in a direc1ion ~ransver~e to the direction ~r rluid flo~ ~ith each l~assing v~rle~.
The fate al whi~h the vortice~ are shed by the bl~lfî b~)dy i~ ,trop()rli~,nal to the flo~rate cr the rl~lid. ~ strain g~uge, rie7oeleclric element. t)r other 10 similar tran~ucer, att~ched to the ~ring creates an ele~tric pul~e ~or each pa~sing vortex. 'rbe re~ulting pl~lse ~ate ro~ms the pl'oportion~l bas~s fo~ thecc,mputation or the rlow rate One ~igniricant and persi~lent r,r~blem in modern vc)rteY ~hed~ing rlowme~r~ has been Ihe inabilily lo eliminate extraneous noi~e signals from vor~e~ rre~uency signals. Sinc~ the vorle~
sensor ~etects ~he mechanical reactjon Or the wing to Ihe action Or vortices, thc ~ing al~o picks up all o~her mechanical action~ including structur~l vibra~ions of pipè lines, low fre~uency acoustical noi~e~ penetratinR acro~
the pipe wall, noi~es ~ssociated with flow rluctl~ations unrelated to Ihe vcrti(~ and the like. ~he noise problem becon~es particularly seriou~s when 20 the blurr body and wing are mounted:on the end of an elongated ~em or probe inserted into the mid-portion of the ~luid ~trea~ in a cantilever mount, and it is f~lrther e~aoerbated ~hen it ha~ ~o detecl a low ~relocity rl~wOr a low density fl~id.
An ultrasonic means ~or detecling the ~ake created by the ~tortes generator plovides one ~ttempted solution to the aforementioned weaknes~

O C: T -- 1 ~--~ 5' T H IJ 1 ~: 3 ~ 2000700 p ~ ~

~ ~ . .
in the pre~nt-day vorles shedding flo~meters. Ho~t~ver. the vc)rlex ~hedding flowmeter ~ith ultrasoni~ ~al~e detsctic)n ha~ its othcr ~e~knesses or it~ o~n ~or exa~ple the error inlroduced by the ~ubbles and parlicle~
suspended in Ihe fluid medium dis~ort the ultra~onic ~ignals. Al~v. th~
ruggedness needed for durability. temperature, and pressure extreme~ in nlan~ practical applications furt.her limits ultrasonic vortel sensors.
~UMM~RY OF rHE INV~NTlON
Accordingly, it i~ a general object of this invention lo provi~e a more effective metho~ and apparatus for ulilii~.ing vortex sheddinR and othe~
1~ fln~rmeters by ~erivin~ ~ mc)re lls~able flo~rate sighaî fr~m ~hich t~
measure flow.
It i~ ~Iso ~ gene~al obiect Or the ptesent inventjon to prnYide a ~,~orteY
or other îlo~Pmeter sy~lem having electronic means lo ren~o~e or minimiJe noifie signflls ~rom the Mcnr Yorte~ frequenc~ signa~s, ~herein said rlow or vorte~ frequen~y signals are used to co~pute the f~ow Anolher genera~ object of the present ~n~ention is tv proYi~e an inSerliOn, probe-mounte~ vortex or other flo~meter system ~hich comprises a naturally hi~h ncise environment, ~ith eleetroni~ ~eans to ~emo~e the noise signaîs from the vorte~ fre~enc~ ~ignal~, ~uch th~l ~he v~rtel or otl,er 20 flow ~fe9uency sign~s can ~e u~ed more e~fectively to comp~te t~e flow.
~ nother object Or the pre~ent invention is to provj~e a vorte~l rlowrneter system having ~ strain g~uge or other appropriate tl~ansducer and inte~ral firm~are capable of e~i~inating noise signals from the vorteY
frequency signals.

O C T ~ ~ T H U 1 7 - 3 C~S
2~007Q0 P . ~ 8 Another oh jee~ or lhe pres~n~ inventi~n i~ to pro~ide a vOI leY
Jl~wmeter ~y~te~ ha~in~ integrsl firnm~are that utili~e~ a c~mbin~tihn Or rrequency dt~m ain conversions ancF cro~ correla~i~nal technique~ to eliminate noke signa1s rrom the vorte~ rrequenc~ signal.
A more speciric object of the present invention is to provide ~, di~ital çi~nal prc~es~ing technique capable Or recognizin~ an~ sorting c ut Ihe rrequency domain pattern Or a vorteY or other Mow transducer signal a~
being diflerent relative tb the pattern of noi~e.
Additional obiect~, ad~ants8es, and novel fealureç o~ ~his invention are ~et Jorth in part in the de~G~iplion lhal ~ollows, and in p~rt ~ill become ap~arent I~ those skilled in the art ~pon examination o~ the followin~
~pecirication or may be learned by the practic~ o~ the inv~ntion. The object~
and advantage~ oî the invention may ~c rea1ized and obtained by Mean~ Or th~ in~rumenta1i~ies and in combinstions particul~r~y pointed ~UI in the appended claim~.
To achieve the roregoing and othel obiects an~d in accordance ~ith the purposes Or the present inventi~n, as embodied ~nd broadly ~eçcrjbed herein, Ihe method and apparatu~ Or thi~ invenljon ~ddreç~es the nc~i~e problem by ~pplying a uni~Ue digitaî signa1 proce~sing program to any v~rtel or other rlow transducer ~ignal. The program u~es a ~mbin~tion of ~requencS~ domain con~ersi()ns ~nd compa~atlve processing to i~olate the transdu~.er signal fro~ the noise signals. T~us, the vorte~ shedding ~iowmeter can p~ovide aCcurate rea~ings in high noiSe environments.
A signi~icant feature of Ihis invention is the disccvery and recognition of a signatulre ~haracteristic ~'averor~ ot the flou~rate ~i~n~l ~s ~' 2C~700 `J

~l~st!slrt from n~is~ s~grl~ls in th~A c~rn~ S~ en~J ~r~ n~ n ut li~n~
it.Al pr~ A- slrlc~ t~Acr!rliqu~C t~ C;~ r~t~ out ~.r,d ~ nin t~ th~ n~lC;~ r~
s~.qiicll ~ir n~ t hav~ t s~n~c~tJur~ ~h~r~ct~ri~tic. In doin~ cO, ~ lc~..r~.~ n~ n~ r ~i c~rr~ , su~h j~t,` pr~Af~ra~ly ln~ of th~ v~r~" sen~r ~ nals ~Y~ t~ n ~v I .. p~ t l~,d of tt2~ ~, SU~h as up t~ , s~Qn~s; . t ~ pr~t~i ~, J c;~ rltJ oll ~S~lm~ .~, irit~r~.~d.l~ ~ ~iiS tim--~ -domam s~rr~ d ~ is th~. t~.Y~ns~^t m~ t.Ct c~ St~q-~n~y-d~ p~ sp~.~t~ n. ~r~7t~Ye p~t.~e~ tC;~ f~qll~s~7) ~!y m~n~ ~!f ~ k~C;t. Fo~ r tran~f~rm ~al~ul~tion. ~rl ~r~er to irnpr~e t~i*
'l;,t~ r~ t~t~o ~s.~ re~ C~ tt~r, S~Y~l-r~ S` C;p~
t,.~k~n ~rorn pri~r s~m~ling perio~ n~ av~2-~cl b~f~r~ fur~l~r ~.i F~ ln~ i~ do~e. T~i~ re~ult~ in the c~Pr~e ,t~ot,~r C~;7~e~ trlln~ J~it7.
f~er~g F~ iery 30w fr~u~ cies ~re th~n fem~c} JrC~I~ t~ ver~
~Sr3 Th~ rec~.ultin~ c~ver~e PSr~ ic; U~en ~roc~-col-re~ d W~J ~ seI-ie~ of t*mplat~s ~elec~d noi~ p~tt~rn~ t have ~nd~ re~r~nt~tlY~ c1i r o~e. rt~is re~ult~ in the n~ r,plat~ c~oss-~orrel~tion. At t,he freq~el~C7 or e~ch pea~ in tll~ no~ mpl~t~ cl-o~-correl~ n, th~ erage P~Lj is ~h~cke~ t~ det~rmin~ ~f it more ~loselv ~pF,ro~ma~ a s~ t~d ch~r~ct~rlsti~ Y~rt~X or f~ow ~gnalJ su~h ~s wide b~nd~dth ~th ~ ~ of ~, ~r ~ noi~e signal, çu~.h ~ narrow bandwidt~ Q o~ 3~, where ~ is a 20 me~sure Qf t~ dEi ~and~i~ of a s~n~ e pe~l~ in tlle a~erag~ Fi~
1~1~.~ rno~e like nois~ n ~e p~ak, ~n~ surroun~ing points- are r~mo~d, s filtering out or removing the no~se compon~nts ~f the si~nal. Tnis s~r,i~z d sign~l results in th~ fil~re~ PSr~.
T~e frequen~y o~ ortice~ is t~n e~ d ~ first cro~-~rrela~ing ~ filb~r~ PS~ Wit~1 a Ser`;~so~ tes- ~ (Sl~na~ ter~lat:es) O C T -- 1 ~--8 ~3 T H U 1 ~ : 3 :~
- Z~ 700 ~ `

v~ c~d ~n~ rhc; ~imil r t~ rt~xol- f~thf~r ~ n~. Th~
re;~ esttJ~-c~t~ ~f the fre~u~n~y 15 t.-he Irequ~ncy ~t w~ t~ n~mum ~ue n th~ ros.c.-~orre1~u~n Th~ e.~m~t~ ç~n ~e fU~ - rf~fi~e'i 13--c~c~ cot r elattnc; ~n~r e~n~l t~mpl~e cen~red ~t t~e estlm~ed tref~urn~y ~ith the fi~ PSI~, repeatin~ this ~rofve~s if n~cesc;~ry. T~e r~ lt. IC ~r. a~ rat~ ~ of t~ er~e trequen~y of the u~lrtices T~l~ p~ ~r at Uie *~titna~d f~uen~y i~ ~hen c~ul~t~d ~r.~.l C~ . rf~ lum. ~ pt~ r 1~ thr~h~ c.~ rt;l-t~ f~ui~ d~n~it.~r ~n~ ~t~n~ fr~u~n~y I~ U~ r li) t.~ lCl~ po~Jer~ tl-le r~ult i~ ~et to 2ero, ~nd t~l~ en~ir~ prc~ r~ t~
r is a~ tni~ th~h~1d I~Jel, t~l~n tl~ troni~ c~utF~ut <~r~:uitry i~ cet t~ output a ~ tle th~t i': pr~p~rtion~1 ~ th~ r~ult~1lt.
e~ t:~d lr~q~l4n~y The fl~wr~t~ i~ dire~.tly cornput~d th~ f~-on..
~R~EF DES~ TIC!N OF T~E ~)k~.WlNC~S
The ~ p~nyln~. drawirlF~ hi~h 2r~ incorp~rat~d in, an~ forrn ~,~rt c~f, the sp~ci~ tion~ illus~ate ~l~ pr~ferred em~odim~nt~ ~f th~
In~..ntion, and t4~..ther witl~ e d~ription ~erve t~ ex~lain th~ prin~ s ol the inv~ntion. In th~ dr~nng~:
Fi~.ure 1 is a sch¢mati~ r~present~ion ~1 ~e l::on~vepts ~ nent p~rts of ~ Yort~x she~dln~, flos~nnet~r;
Figure 2 is a ~enlati~ ~iew showirlg an ins~rt~n tyF~e ~ort~x sh~dding llowmet~r ~nsor in ~le ~nt~r of ~ pipelirJe, Fi~-lre 3 i~ ~ sid~ elev~tioî~ iew Or ~n lns~rtion-typ~ ~ort~a~
sheddin~ fiow~net~r ~ it i~ typically moun~ in a p1pelin~ ~rith the pip~lin~
sho~m in cros~ ~ection;

O C T ~ ~ T H U 1 7': 3 ~
-- ZC~70~ P . 1 el Fl~ 4~ fr~ Y-~ti~nc~ n i~-~s~t-ti~r~-ty~ t-te~
t~ i2l~ f~ .A~ t~ el,~or a~p~t~c i~ s~ rl Ir~ ~g1!~e 4 Fl~ur~ 4~ s ~ sid~ ~If~ ttion~tl ~e~ J c~f t~ yort~ clJe~ inc~.
fl~wmetkr sf~ns~r in Fi~ure 4;
Fluur~ 4(c) is R ~c~n~ituclin~tl s~t~tic~n ~?f the ce~ls~r ln F~ re 4i~;~3 ~ t'~p~r~ t~y~ ~f th~ ser,sor ~m t~ken ~ong Sln~ 4~-~c of Fit~r~ 4.
~ i~t~r~ 6~1? i~ n~c~itl~din~1 ~e~ rl ~f th~ s*ns~ in P~ lre ~ t~.~n ~lf.~n~ lin¢ ~f~l 4t~ ~f Fi~ul~ ~b sh~w~ s~Y~ c~
Fl~llr~ 4~.ei is ~ ~ot~m CfOSS C~:ecti~"l~f t~e s~n~ r ~rl rl31!5~ -t~ r~ line 4e-4~ of Fi~,ur~
F1,~ure 4~f~ is ~ tt~m ~r~ ti~nal ~ .nco:- in Fs~.ur~
t~ken ~l~rlg l~n~ 4~ ~f ~i~ure 4b ~howin~ -ain ~ug~, Fi~ure 5~ an~il 5~ et~er illust~-~ in s~ m~ nO~! t~e ~ort.~.
s~nSor ~roduces ~.n an~log ou~,utJ ~ith F~w~e J~t r~F~reserltin~ ~ slde vle~ csfthe ~luff b~dy and ~e sen~or ~ ng ~nd Fi~ re 5~) r~preC;aritin~ t~:~ sros~t.
vi~ c,f t~ r~tificr ~n~ and t~e st~air~ ~a~g~ n~ducer ~ctnne~t~ed t~
~tentiotnet~r~nd . ut.puttin~n ~fialo~ nal;
Fi~ures ~a~, 6(b~, and ~c~ ~re ~r~phs sh~ing U~e rel~f~n~ p ~tw~n time d~maln ~nd frequency d~m~in sign~ls, ~0 Figure 7 is a ~t~p~ illust~tin~ the concept of band~dth and si~nal sl~rlatu2 es, Figures ~ (b), an~ ~(c) ~re ~r~phs illust~tin~ the ~onceF,t of ct-~;s-correl~tions;
F~ r~ g is ~ flow dia~ram of t~l~ noise remo~ ps ~or~ng ~ t~e ~,~e~nt inYen~ion, O C T -- 1 ~ --8 5' T H U 1 7 : :3 3 2~ 70~ P. 1 1 r~ fl~J dl~.r~m ~f th~ nal rfJ~as~r~ ont ~t~ps ~f t1 elP~t! ~n~ r~c~ct!t ~.c~ording ~ t~ pr~C,~nt ir,~ntiorl~
~l~r~ ! 1 ic ~. flo~r dia~f~m ~ ect~oni~ ç~mpen~nt~ ilJ t~e ~trc~ni~ p~c~C.or o~ t~e pre~rJt inY~ntion, Fig~re 1~ is ~ ~r~ph s~,o~,ir~ ~ r~F~res~nt~iye r~w freq~lency d''I
n~l fron~ ~ Yort...x s~ns~r;
Flg~ g,r~ph sho-~in~ th~ ~Yera~ -al re~,r~ent~t~
r ~J frP~u~r~ ~y domain si~nal~ lik~ t-h~t o~ u~
P~.r~ p2~ sh~ g th~ '`f Fig~lJ~ 1~ df~' ~ ~0 2~ r~ 0t~ . YOtt~X Si~fial ~ign~tut~;
~i~u~-~ 15 is ~ ptl ShC?Wln~ t~ sign~t of Fi~u~ f~f ~ s~c;on~
cr~s~-~ot r el~ n.
F~ure ! 6 1~ ~ graF~h sh~ng a VoI t~X t~mpl~
Fi~ure 17 is a gr~p~l showin~ a noise temF,late;
Fi~ur~ l& i~; a gr~.ph sl~?win~ ~ ~ombir~tion YOt~ d noise Sl~
corr,p~r~l t~ r~s of nois~ t~mpl~es;
~ i~ure 1'-~ is ~ ~raph ~ho~ing result~nt signal cleri~e~ Irorn t~e cros~-correl~tic~n of Fi~ure 1~;
Figure 20 is a gr~ph sh~wing the i~lat~d ~ort~x si~nal ~om~red to ~0 c;~ri~s ~f ~,~02t~x t~mplates; and Fi~ r~ 21 IS a gr~ph ~ nrlg th~ ex~ct lo~tiorl of ~e pe~k c,f t~
Y~r~ex si~nal W~IIG~ result~ fron ~1~ cross-corre~ation o~ F1~ure ~0.
rlETAlLED DES~:~IPTION OF THB t~EFE~E~ El~ IMEllT
T~e si~nat i~erltifi~tion ~n~ p~o~e.;o.ing t~chniq-~es of U~is inY~ntic)n can ~ us~ for ~i~nals detlYe~ Ir~m any ~ ~ numb~l clf fk,w 1n~t~r typ~

O C T -- 1 Z -- ~ ~ T H U 1 ~: 3 ~ Z6~070C) p . 1 z nl~ t~ s~ ucfr~ t~t cen~ el~c:t.~o~C ~ Isc~!~ t~ t ~ e ~ liC~tiV~ c~,~?
rat.~ ld m~c;~:~d itl ~?"~-c~n~ou~ t~ ,iC~n;~.lc ~1 p-.lrp~ t.~
C1~r1~jOIIJ ~ t n~t ~ 1imit~ ,n ~UCh si~n~l~ prr,du~ y ~r. in~.~rt~c!r, V~rt~x ~P~.ding ~1~ m~2 will ~-,e ~ ri~ ec~u~ t~ley ~.r~ p~t-ticul~rl~
bi~Ct, t~ ro~ r~-SC
In ~c~c~r~n tl~s~ ~Ihr~ de~cribe~ ~urp~ Ut rlC!t îf~r l~.rnitlnC, t~ pli~.tiC?I3 ~)f thiC in~ntior ~r~J ~ny p~rti~ r flc~r I~.'et~l tV~ r tr~n~ c:c~r, t~ r~r~ tt in~f~n~ ll C:~l Utii1Zt~ ctl. i~.t~t~ IUf~ tr~i7 c~n~i ~rt~ - ac-c~f~rlbl, ~rn~eiIi t.h~ ~n~:~ C~r~ t- le~f~t~ th~ tA~J
1~ cre~t~l t~t;. t~-le blulf ~ y. Tn~ lf~ lt~tnt ~ sn~tin~; lilt f~tr~ c~ ~ID
~rt~ *r:~r~t~ ,uenf~y ~ ; senC;~S~ ~y t~ in~ C;etiS;~ }l F~tc~d~ ,ron2c si~-n~ls proF~ n~l irJ &rnplit,u~ the pî~C~r~ p~ k~, ~nd irl IrequJ ~ ies in~ic~tive ~f, t~e ~f~rU~ lS ~ 7lS t1'lf~ ext~-~ner.~u~ n~f~s in ~ e sy~t~m Th~ sign~ r~ arnplif~ed, ~rJ el~ctrç~ric~lly filt~ t!~
r~m~J~ hl~h fr~ en~y ~r~tent, sin~ th~ ~rort~ fl~ nal~ rel~tl~e!~
requen~y ~he si~ ich in~lu~es ~th vort~xi~rld reln~inin t~
~requen~y nois~ rrlr~nent~, isi ~rie~ pref~ri~t~ly t~ ~ flrrn~ e cir~uit re noise ~ign~ are se~i~ri~t.-~ frc,m the ~or~ ~r~u~ncy ~i~n~t~C. b~-dlglt~l si~nal ~rocessing t~chniqu~s. The me?~sure~ fr~quency ~ut.put of th~
~0 vensor 1~; Upd~t~ rio~lcally, sucn a~i onc~ e~ry ten se~o~lds, i~s eit.he, ~tlrrent (4-~O~A~ or ~requenc~ 2~ Sig~ llr~t~y corres~-~nding t~ tn~
fl~. A fin~.l c~lnpu~tion of fl~w is th~n made.
~ross ~c~rrel~tion~ hni~lue~ are wi~ely used In t3~e ~rt of di~it~l si~ al proc~ss~n~ to ~enote the siimil~riti~S of t~ w~ f~Jrm~;. I
~pe~ifl~llv, t~le simi~arities ~etw~en t.~o ~eform~ ~s f~urld by ~.umrnJng g o r~ T ~ Z --8 ~ T H U 1 7: 3 ~
- P . 1 3 2C~0~

t~ S~~ ed ~ro;l~lcts ~,~ e~ch ~rr~ for~ ln~ti~!n~ .r.~-r~lr~t~fJI-~ c~-l b~
t~ ;ht ~f c~s ~ ~r!~t,~ , Up <~f ~ f~ m ¢otn~n~nt~; o~ cil-~Jil~rit.~,~ t~.t.
t.W~II W;~Ç?fOlniC. As ~ed ~rein, ~lo~ ~o~rel~ti~n i~ u~.f-d t~ ~t~rm~*
~hPther th~ ~n~ sh~ of ~ ~ort~x freclu~ncy s~ atul~e is ~ en ir~ t~
r,r~ e'X. ~r~.~ef~rt~s ~f ~ s~n~ a~in~ bot~ Yort~X ~nd noiC~ c~?r~l~on~rJtc Tll~ flrc~;t ~t~p m~t~,em~tic~lly ~ 4n~ert ~y n~P.r~r,s ~f FFT t~
in~ ~X ~r~d n~-e ~Y~f~rl~ fron~ d~lm~.in t~? ~1 r~n r",y ~m ~
Th~ c4m~ine~ ~dt7e~rrn is no~ ct~re~ in fr~qu~n~,~ d~rr~ n p~t,t~rn.
10 The c~mt?ine~ ~Y~orm ic~ en multir~libd b~ del~y~d ver~iet~ .n~
n~rr¢~ ban~ th 1iOi~ yeforrn~. Th~ s-e~;ultlf:~ cro~-c.c~l-r~ tic!n ~a~ef(lrm ~rlt~in~ ~nlv t~l~v~ f~-equ~n~y ~ompon~nt~ ~orl~t-~lon t~ th ~.~r~7ef~rmc. T~lus, the noi~ qe~rm c~mponerlt c~n ~ ulle~ ~r~m th~
il~mt~led ~sh o mlnnation ~rt~x an~ noi~e ~.~eforr~.
bne t,~F)e of cro~ frel~ n mathem~ti~ may u~e the dbl~iy~
e~rl-rl of the san~ v~form %s ~ ~roduct. T~lis produ~t is calle~ ~n ~utocc)rrel~tiorl When the aut~orrelaU~n is pr~sed ~y tn~rJs o~ ~n FFT, t~le r~ lt 1S called ~ power ~pe~ l d~c~.ity ~PS~.
P~eSerrin~ first ~ ~igure 1, a fluid sta e~m in ~ pipe ~not showl~ in ~0 Fi~ure 1) flo~ in the directiQrl or the arrow 44. A t)lufS b~dy 1 t~ posit~one~
in U~ flOW stre~ e bhlff t~dy 1 acts ~s a i~ow t)~r~ r ~nd h~s ~
t~ndency to cr~ate YoI-tices 2, ~, ~, 5, 6, ~ do~tre~m th~reSrotn. The t~ortice~ 2, 3, 4, 5, ~, 7 ~re ~ d Iro~n t~e bl~ y in ~ rn~ting f~shion fr~m si~s 1 1~ an~ 120.

- 10- .

ZCI~)Q70~ ; P . 1 ~
~J

Pc~r ~ C:f.`n5t~rlt~ ?~', th~ t~ C3i~rf~r~ti~ c; ~.t~ er~ rtice~ ~ throut~h 7 ~r~ n~t. f~X~t~y ~ ~C~1n~. Ho~tf~!er, t~l~ tirlle ~i~f~rer~ lc; ~ t.~ cn t~l~
~ort~e~ 2 thro~~ 7 t:~1J b~ r~:ed m~ airly ~ ural* ~ir,tit.y of C~ r l~rl;t tjm~ ~C~I ~rlY C:~1n~;t2~11t ~ J r~ For ~ ty~li n~ t. ~ e.i~t,' f~,ur YC~I-tiC'eS p~r ~v~d ~f~ ~ert~?. T~* fl~w rc~t~ ol flui~ l~r t~lle; 6~ ~iel t2 YOl t~ freq~l~t~ n ~b' ~om~ute~;i fr~m t~ following ul~
K'I ~V~-tv~ ~r~u~ncy in Hert2 IJbr~t~d ~ lin~ c~rl~t;Jr:~
. m~t~lod rrlu~t b~ pro~id~ e;~ns~ q)~rltlt.y o~ ~ortic~. C~ jt-~
~ n~ Ul~ ~!u~ 1. F~!- F~ -pf~ f ~les~;t-ibilJg t~lic ~n\~t~t~ , al~
el~tJ~t~d, ~l~nd~r ll~ing g is S~l~w~ -, pl~ ie~ ir~ ure 1 pl~d lr~ ~,e 1 e c,! ~Oî~ S d~nst~ m Irom t~& bl~f~ body 1. The ~ortices ~, ,~7~h h~r~ a hi~.h pre~u~ t,~r ~all 2-, g~, 4, 5~, 6~, 7, r~srlectiY~ly.
ec~ hi~l presc.ur~ out~r w~lls ~', 3', ~, 5', 6, 7 move or ~i~r~te t~Je wln~ 9 ~ro~ sld~ t~ sicle ~s indi~ d t~y arro~ 4~ ~ thçy p~ y ~he ~ h ~ressur~ out~r ~11~ ~', 3', ~', 5', 6', 7' mo~e t~e wing g in dir~ pr~,~rt~ic,r,~o t3~eir rel~tive stl-ength~. Addlti~n~lly~ t~le ~rtices ~..ff4, 5~ 7~mo~
the wing 9 on~e ~or e~ch vortex p~in~. by. Th~re~re, it is po~lt~l~ t~
~0 m~ure ~)t~ t~Je st~-en,th c~ e vort~c~C ~nd t.~le freq~lenc~ ~f the vc!rticee;.
The present ~nvent~on relie~ prim~rily on ~e freq-~ency of ~ ttice~
meas-~t~ ~lo~-ate, t~ut the s~eng~ c~ o~tlce f~rces on the ~in~ 9 ie.;
u~;ed ~s a credibility ch~c.k. A strain g~ge ~not show~l in Fi~ure 1, but sn~m in Figure 4~d~) or some ot~er a~proprj~ n~.ducer d~ e, suc.~ 3S ~1 piezoele~tric elefnent ~r.~t sho~m) ~n be ~t~ h~d or c-,nnected t~ the w~n~ ~

O C T ~ ~ g T H U 1 P . 1 Z~ O' '~

t~ r~ t~ ct~nic ~!t~ S ~ 1;?~ t~ ql~r~c:y ~n~ t! ~SI~ t.~e c~e~f~lin~ ic;~c~.
in thP typic.~l operatirl~ enYironm~llt~ o~ a ~o~ h~ n~ Ilo~ n~t~
t,~ re ~ t m~ny ~ nP~us lorc~;~ ~n th~ flo~i sts-~am, ~ ~e~ d ~y nolC;~
w~ n Figut-c~ 1 The~e noise ~ 10, as wPIl ac m~han~f.~]
Y5tiC~ n th~ p~p~ tranc;l~nitt~d thro~h ~ ~in~ un~!~,J ctr~)¢tur~
e t~ c ~i Such noi~i~ w~ ! al~ n~?t-at.~ rltn~ y ~!t,~
C'!]tlPly ~.tr~-~;hed t~ pip~lirl~ c~r~yln~ ç ~ re~m ~n~f~ bu ~rl~. tul-buler~ce ir~ Q~t c.t,l-eam Th~ r!~ s l ~ arld ~t~.le~~ ~tJ ~l-t~.~t lf~ .~it~r~ti~ns comtlirJ~ ~it~ the YOf~ , 5, 6, 7 t~ pr~ f~ r~iple~
~!~;t~1ni~ sign~ b~.e~ orl t~ moYemerlt~; ~f win~ g, suGh as t~t- C;~ n f~"
t-~ Fi~ h~ p~erlt i~lY~nti~n p~-o~id~ a rf~
~e~r~t~l~g and ~limln~in~ th~ nois~ nals fr~rll the ~ It~X ~is~ ; ir~ et k~ o~t~in ~ r ~ ~tex sl~nals fl-~m WtliÇ~ to m~sure rIc~n-2~
~ le ~luff ~y~r~ ~ss~m~)ly 2~ i.c. sh~wn in Fi~u~-e ~ sup~ort.~ in th~ mi~ f a pipeline 21 by t~ stem or pro~o 2~. Thi~ ty~e of ~luff ~cly~ in~ as~em~ly 20 and pfob~ moun~.ing is re~erred t4 ~ an ~nse~ti~n ~YF~ fkJwtn~ter. The ~ss~mb1y ~0 c~n ~e ins~1t~.~ into an~ rem~ ed rr~m ~rl exisUng iar~e pipe ~1, such ~s through 2 ~at~ 'Y~lve ~ody as ~nll be ci~cribed in m~re d~t~ w, without disassemblir~g ~he piF~ er embc~ir~eIIts of ~uff k~yf~nng ~sset-~ s ~not S~JOWn) ~ret~h a~ro~ ~le ent~r~ dl~Jet~r ~f tb~ ~ip.. and are mount~d in hou~ings ~r cou~lint,s ~osi~on~d ~et~een, ~nd (as~n~ to, t~ ~ections oI the pip~ . The 12~t.~r typ~ ~f ~ r~
re~rred ~o as in-line ~ort~ flowmet~r~. T~le appar~us and meU~ods ~ ~,is ntion can b~ used ~o~ si~.n2lls pr~u~ed ~y bot~ the in-l~n~ an~ inserti~n O C T -- 1 2 --8 ~ T H U 1 ~ : 3 5 2~Q~70 nC~t~J~ t,t~ on~~ e~ ,r~nul~ ny~ would l',t',b di~f~
S ~ppro.!rl~tC! t~ tlot ~ th'~t t~ rn~th~s 1~r d~t~rn~n~nc~. Ule ~ ~GI~St~r,l.s s; lnst~ ?~tl~lrlc ~ f~ ter~ w~ nwn ill t~jc; fi~ r!1 r~ t t-~ e~t~er~ f~ .n ~ncler~rl~in~, o~ UliC; in~e~ ti.
7t t~J~ L ~rl~t-~ t~t ~ t~ c~ ~n~ t~ c!nc;~ t f~.
t!~ rt~c~rl ty~ ,nn~ u( h ~c~ th~ d~"~t~ n ~i~UI
f~ t~rc ~ cnr~!F,~ r~C,f th~ f~ fl~w V~ it.Y f!~m t)lc~ m:~Ylll~ If~ fl~J~`.' .~el~ . W~liC~ ne~ t.. he c~ r c~f U,~ p-p~linP ~ ,f~
~ it.~ c~ ?ly ~ Th~ ~h~ f ~,~ fl~id ~ J ~ f~ ' liff~r~ ff`~ ri~u~ nd ~ c ~ met~ C.. ~ iff9'r~nC-f~' c~r~ e~t~
i~t.~ U~ rJr~f~f nt. ~r vent.ic,n ~y th~ uC;~r, pr~er~ y ~y us~ n~ ~n in~ t lXI~Ut~
4 ~ lC~ I in Figure I 1.
~ f~ierritlg next t~ Figure ;~, ~n ln~ertion-t.~ vo~x f~owrflf~ 3f~ C;
sh~ nount~d in a pipellJle 21. The ~luff ~y~ n~ sen~or assemhl7 ~
s~ tporie~i in th~ ~nt~r ~ pipeliil~- ~1 by ~he s~n~or s~m ~. F~na,e~ ~ij4 .nd ~5 su~rtc~n i~ tion ~7al~ 36. T~le isol21tion valYe ~ ! can ~ a ~t~
~r~ f~ ~nd ca~lows removal of the in~rtion t~p~ ~ortex flo~-nete~ for maint~nctnc~.
~. ~econ~ lation valve 37 p~tmits ttl~ ~pt~onal addition of ~ pressl!r~.
20 tr~nsduc~r ~6 t~ e in~ertibn type Y~r~ flow~net~r 30 h thre~Pd mc~untin~ assem~lly 3~, ~,g ~llows th~ h~.ndiP 3~ t4 accul~tely p~itior I.~n~r~ln t~le c~nter ~f the pipelin~ ~ W~r~. ~n~ e sensor ~lr t~an~mit t~le si~n~ls ~rom the sen~or ~ tb the ~l~ctronic pr~ces~or 4~.
The ele~ nic pr~cessor 40 ccnt~ins 311 t~le cir~uitry nec~C~ry t,~) ~x~ut~ Ule ~r~s~r~ in~r~tion. ~ri~le ug~r t~libr~tion p~rarneters nl~y 1~ 1_ T~ 13 ~ T H 1~ 3 ~
P . 1 ~
2~ )Q
.~_,, .

.fJ ~ t~ int~ tr~ r~ t~ r~ ne. ~ tl~t.
shoW~ , . cQn~Y~e~lti~nc,1 2n~nn~r P.~ft lrilla t~ F~ r~ 3 ~nd ~.b), t~nc.~r Ct.A!II 22;sl~F~r ,d~ .ri~ ss~ 1y f7Q. ~ ~y~ind~-ic~l fl~t.A~ c~ 4~ ndit~ t~l~ fi~
c~l-our,~ J~ blu~f ~odyfwin~O ~ nbls~ 20 ~n~ o!~ r~A,~n ~
tllr~!ulen~ Th~ lon~itlldm~l c;~¢t.ion~ w c)f ~Iur . t.~ w tr~m lef.' t~ ri~lt, ~e; in~ d ~,y th~ ~r~ w G4 T~1 ~ 45 ~rPatR~
t~ vo1-ti~e~. ~c; ~s~rlb~ Th~ f~ r~ C r~t th~ ~tJ~ 48!
l~ ~c; t~ le~ -;y ;t~r~ ik~l~t~ ,J ¢.~ e i~
~0 ~li~ t~y ~r Y~ R fI42n c~ -c.:~de ~ scri~*A ~b~
lhe ~7in~ 4~ ir~ ~3i~ ~rr~ -r~ t., ~lc; ~t~ ; th~fi ~?C~ 7?.~i~ 1 IC
~tu~ hion~d c~ut o~ ~ rn~,n~ , urlit ry t-ll*t;~ ~ t}
5t~m 41 ~hes~ eof th~vai~eor ~n~ mic~ is prefer~ J a b~y in t~Je~ ~
forrn ~f ~ fixed fr~ be~m ex~n~n~ t~t~en ends fi~ cl t~ ~r:,~r ~rJc~ r ~i !OSi~y r.,osti~ns 51~ 5~J fe~pe~iyely~ n t~C;t be s~en in ~1~ures 4(~) 4~d), arl~
4~ free ~ s 54, 55 ~f win~ ~ ar~ s~p~r~t~ ffom ~h~ b~-~ 51 ~7 narro~ cut~ 57~ 5~, res~p~ti~ely. A s~-ain ~auge 4~ or ~ p~ir c!f st~-~irl g~e~
t~ d~ t. l~i-direction~l moY~rnent, is 5ho~n in Fi~,-ur~ 4(f) mo~nted t~ t~
~tt~m oI the w~ng 4~ in ~ re~ 50. ~rhe t~-~n~rnitt~l ~rir~s 47 from the 2Cl st,rain ga~e~ 4g run t~lr~u~b a ~on~uit.- 46 e~n~in~ upwar~ly through t~ r~
~ent~r ~f t~e ~lu~f bo~y 45.
F~ r 5 d~p~c~. ~e ~quen~e of how t~ ff b~dv ~5 an~ ~n~ 4 produ~;e th~ ~rl~l~ si~n~] 5~) s~ich is indi~ativ~ o~ ~e p~in~ ~ OrtJ~S ~
noi~e ~r~Je~ The wing 4~ flex~ from si~e-~o-sid~ ~s s~lowrl ~ rf~ ~S~ j A bC- power s~lpply 300 pow~rs ~ p~n~iome~r 400 ~*lich sense~ t~

O C T -- 1 -2 --8 '3 T H U 1 ~ ~ P . 1 3 Z~7~)0 , -~c~ist~ ce ~ t~ t~ r~ u~ T!~ reC;is~;tn~ ¢ll~ng~b~; C!~:cu ~,t: ~ s it-~m c:i~ ¢~e T~ F~ t~t,~ .n~t~r ~ t~ t.~. ~r, ~n~ C~ rl~1 5~.~r~ p~ tfJ, ~rld inc~ic~t~ f, t~ mc~ o~ th~
~11'1¢~
Fic~:ure ~ lu~t~-~ies ~n ex~rnpl~ i~e~l ~n~ ic~r~ ! Irom Fig~lr*
e, ~, it t~ h~ lbOk if th~tY~ ~St~'~ n~ nois~ ~n~ e~ch V7~)f tf-X *~C~ s~lef~l exc,~tl ,J
1 ~r~ l ~;'bCf.'J!~ t Th~t ~tnplit~d~ ~c~ri~ pl~ nc3 m~ c;~ t'r!~' *~ .~r~ s~l~d al~-natR~y lron~ e~ l lo an~ f t~ u~f C~e~nl 2n Fi~ut ~ ,us~ sc~m~wh~t n~r~ r ~listi~
lf~ re~ t~ tion c!f t~ sligh~ly ~2-ying inCt~t~t~n~uc; fr~q-l~îlc~s of v~rti~
~31.~t~t~ c~.tn~t~î-lt. flc~ r~t~ c-e irl re~ y t~e ~ rti~s ar~ n~t sll~d c~
ex~ e ~.~r~ tim~ interYal~. Hnw~Yer, ~ dis~ ab~v~, the ~v~ra~
fre~,u~ncv ~Yer f~irly short tilr.e irlt~rYal~ do~ r~main quit~ c~n~ant ~r ~n fl~r r~t~ o~ a ~l~en ~lu~.
Th~ tim~ ~otrl~in ~i~n~l ~f 6~ n ~e c~n~ ted ~I~bm ~he tim~.
d~rna~n to t~lb fr~quency ~Qmain ~ rne~ns Q~ F~uri~r ~naly~ic; in ~ manner nown t~ p~tS~)t~S skil1ed ir; this li~ld, c.o eu~l con~ersiorl t~c~lrljqu~s ~re not descri~ed h~rein. Ho~ Y~r, for F~urpbC;e~ of t~liS 1nYetltiQn, two Cl~
f~ ar~ ~ho~-l in U~e fre~uency dom~in plot of t~e si~.nal in Figure ~c).
20 Pir~t, the fl-e~uen~ies ol ~igure 6~) c~n b~ seen to ~erag~ about 55 ~rti~es p~r ~e~n.~ or~d, th~ sign~l sign~.ture ~f Fi~u- ~c~ ~n tJ~ ceen t4 h~e ~
~ro~l b~n~i~idt~. Th~ fir~.t f~t U~t ~crt~r. 1~ Pq~n~Y ~an be ~v~ra t~ 3 const~nt. fr~qt~ncy ~j5 Her~æ in F~ur~ c~ all~ u~ ~f ~or~ul~
Flow - K x Fr~q~l~n~y, ~s ~ usC~e~ a~v~ onl~ut~ ~I<,w. Th~ ~e~ond ~Ct th~t the vort~x ~re~ncy in Fi~ure ~(c~, i.e., flo~ r~ fJeqllency as pro~uc-e~

.. ~ .. . . ....... _. . _ _ _ . . .. . . . . ..

Cl C T ~ ~ T H 1~ 3 5~ P . i ~
~; 2~ 700 ~, rt~-c;hf.?~ luS~ d~?, h-l~ ~ ch~r~t.~r~crAf b~d ~A~dWIdt~
F~vid~?~; ~. r@cf~ n~t~.~r~ ths.t i.- dic.tin~t frhm ts!r.~iC~I n~cæ clS~nal~-; in th~ fr~?q~ c~ d~m~ ThiC. ~h~r~t~rl~tic sl~n~tur~ ?r W~e ~f~rte~ n~r~t.~d S;1~r~ F~ro~?~d~ t-l ~ ol th~ ~F~rllç~ c~f m~ m~.t~f~' t~mpl~tes ~o 1d~orlti~y Y~l~tex ~erler~t~d c~ n~l~; fr~n-: .nois~: sign~.lc; ~x~r~iiJ ~ lS.
t~ tJt~
S~ J1~^~lly, ~ if~¢~t fe~tu~ -; i2~ er~ti~n is ~ C~ c t ~2~x ~ner~t~J fr~qu~ncl,~ dc~ln~ n~.ls ~ t-~ r(1~t.c; ~-~n~
cll~r~f~;t~ ~ic~r~tur~ s n~ ~t1~ it~!ln ~ ? f~*~ e~
1~ ~7i~it-~it~ c~f t~ t~x ~eT~c1t~d si~llal~ ee~-! S~ n~ n~
experirnerlt~t~on c~nfl ~rlhl~sis t~ h~:Y~ n~rr~w ~ dwidt~C ~ d cur~
c~nd!~r ~ ~om; ~ t~ k of 3~p~titi~f~ r ~ 2-C~ ;iC,n~t1.1r~.
c'1rf~ ~"ke~ ~rl~or r~rf~on-l. Th~ pr~ rlt i~rlti~rl utili7~ th~;f~ sp~
n~iC;~ ~nf~ ~or~x ~enera~d si~t,al sltn~ture~ ~A.~iUl dig~ n;~ rc~f~C;sln~
;~lni~ f.! C!~p;~r2~.t~ ~,r pull U~e llois~ si~ out of the ~urt~x s~ cl]S~
Ulere~!y le~ir~. a ~pe~ ally id~r~tifif~d, t~hrt~ nerate~ si~n~l ~.ir~ r t~ t~""S
th~t showl~ lrl Fi~ur~ j for u~e ~s th9 ba~is o~ f~ow measurem bnt.
For ex~r~pl~ u~e ~8 ~I!u~trat~ tcoulci Pe a typlc~ s~gna~
deri~ Irom t~rain ~ )4~ to the ~o~rnent o~ in ?
2 ~ a~ l co~ ion~; of ~ior~tPx ~ w plus ~xtr~n~u~ noi~ ar d syst.~m v1~!ra~on;;) c re t~e ~.ign~ 1 h~1'; only h~ h e~ n~ls~ frequ~n~i~e filter~ ~ut ~le~t~c~n~c-~lly and tA~r~ it ha~ bebn c.c~n~er~ Irom til~ ~C~îrJ~ jn tc~
frequ~n~t,~ ~c)m~in t~y ~,pli~tion or Four~r ~nalysis ~y n~thbm t~p~ris~lls and cli~ l t~hn~quP~ t wil~ ~ cl~ rj~ed in rnore cie lo~, ~. n~is~ si~nal t~ plat~, Sucn ~ t illust~ d in Figur~ 17, ~, O C_T -- 1 ~ --~ ~ T H U 1 ~ ~

~ 2~ 7 collparcd~tV~ O~ if~ e~es~t~e~c~ua~ enc~
t~a~ Fi~ t~ c~ trlp~ t~ "~ t.o 17~ o~ F~ e ~7 t~d ~*.~ 3C}, ~ to Q ~c t~J~ Ine~c~.lr~ fJI ~ d~ nc~ t~
¢f ~1 iiC~.tl C f~tl 7 t~l~ r~t~.; c?f t~ c.~rlt~r fr~uer~ c~ n~ r t~ nr l~t~ t~ w~ th ~t c~rt~ f fr~ mutn ~ F~lit~ 'rht~ e ~ o~
C; ~1~,?~ f~ f~ t.¢, ~?~ . C~l.>!r~rO.lr~1t.r~ t'~l f~ ti'~f~ ~ppr~-~mât~ n ~i n~
~iC.~ l ChC~r~t~J~.ti~ ft,r ~ c:o~ t~ f~t~ ?r~t~ 5~ e 7 )'1t -.; n~C~ t~mp]~.t~. l 'f l is ?r.,~ . c-~r1~- I s.~ "., 18~ ? ~ t .:I.;t.U~c;.l C;l~r~ 1 C?.t s~ fre~u~n~J c~n~r pcin~; f l. f~ " i4. f, ~IC. L~y Il~ t~ r~ t~ c;lo~-.-c~ t~rt t-~lr~ u~C ~ n t~ nolC;~ tc~mt~ t~c ~2 .,.Y 1~" l~lf! ~nc~ .u~ si~n~ 1, ~e p~ tu~1 n~?~
c-iC~ 1C. ~r~ rlti~ r c~t 1~3.-;t su~;p~t~?~ .<~S th~e r~ l!r~erlt~ f~-f~
nlp1e f~ IO f~, 3n~ f~. Furt~l~r cr~ c;~ -ell~tion ~on5Flut~ n~
e~ h ~ t f~, f~ 7, ~8 ~r~ made aaain~ nois~ ~etr~ t~ 17~ (!f Fi~ur~ 17 ~t~d ~ Y~ eX t~mF~ f ~i~ur~ . 1t can ~e f~1~rmin~d ~y o~npai isor~ îrll ~e~ks mclre ~1os~ly re~emb1~ n~is~ iF~ri~l C(~mpc~rl~rltc;
4r~ n~r~ cl~)c~ rec.~m~le ~ si~n~1 ~om~-oneilt~ F~r eX~ c c~mp~rlisorldes~r~ Qdab~ m}~ td~t~rmln~ tF~e~satf~,f~.~nd~, m~r~ c~c;ely r~mbl~ noiæ ~igna1~. Th~s~ noi~e F~k~ ~t i~, i7, ~n~ ~fi c~r~
~0 t~ n ~ r~mc~veci A ~rY ~1~s~ c~F~r~m~ n ~ s~gn~ 7 ~
nc~ c~ F~on~n~ IS t~n r~,rf.~uced m~ m~if~ J such ~s t~ t c~ in Fi~ure ~. Th~ fl~ rl~içs f 1, l2, ~3. f4, f5, e~., at which the noi~.e t~n~
171 i~ a~lplie!~ is s~mewh~t ~r~ ry, t~ut sh~uld be iJ~ ~ r~rC~ re t~
act.uat volt~ rJ~rat~d fr~quon~y is ex~ted, in ~ eno~$}l int~ t~
accur~ y id~r~tify a~tu~l rlolSe signal F~k~ ut nc>t f ~ se ~ t~

r C T-- t ~--8 ~ T H u P . :~
Z~70 [) ~-;.

e~-lrd~n '~ r~ C;ls~ In*~it.r~f t~ k~ m~ t~eces~arl3y U! rr,~rsor~,~
~et~ ttex t~ p!.~t~ h ~c; t31.~t C~'lf~ n F~ .J-e ~`' iS
~r~r~ t ~aric~u~ sf.,~l"ct~ P~O,l~mie~ uch ~ f IC~, ~11 f lZ t~t~ ~! tlJC' r~ nt. C~lf~ 7 'n Fi~ 2-~ 2~ ~hi~ rt~: t~tnp!~t~ 161 ~s c~!nStt-l~c~
t~m~ uch ~ .~hif;~ n~r~ s~p~ t.~ t "~
r~f.~tf~ e,;tiC. ~2~ d ~ an~ A,t,l-l t~lg~ .tUrC~ e~ t~u~ oî~t~ex j~er ert~te~,. .t~-r~ domainslgn~ heIIf! f11,i~2,et.~,fr~ u~ . c.~
It~:' t~ plate ~ lS appliet~ t~ t~t? CU~ g7 hte ~I~;o ~ho.~n s~n~e~ t.
r~;t~rll~ ut ~ey ~o~ cl~e t~t~l~r e~u~h t~ ~-tevtfi~lelJt.J7 ~IC'f~r~t~ cr~-~orrel~ n r~ults ~ b~ ust~lul, y~ no~ 5~ ~los~ ~.o~rt~J~?r a~ t~!c~ rt~urtde di~ltal pr ~ing ~qui~ ent. ~r t~ bo~)mf~. urln~e~ ril c-lrn~!er~ome ~y m~ em~tisal ~tos~-c~l-relation t~c~lni~u~ t.,etweell t.~,~
~!rt~ t~mpl~t.P~ 1<.~3 ~nd th~ sl~r~ 7, ~ fai~J ~c.
~ rc)~ r~ n of the ~tu~l pe~ , ne., ~;ent~r freguer!¢~,~ f~ the vc~r~
~netc~t~ slgn~l, can ~ ic3erltiri~ r~ul~.nt Volt~X g~ner~t~-~ C~
19~ ~.ti~.h a now known and iden~i~ied pe~ 5 an~ ~ent~r fr~ erlc~ f;.) cc n ~e r~ resent~d as ~.hown in Fi~ure 21.
Th~re are a number ol ~ddition~l pr~f~rr¢c~ ~, ,n~ramet~r~, ~nd ;~0 Ct-O'S shecî~.s th~t. ~ e d~cri-Qed in more ~ lo~ Ho~rett~r ~ e tt y th~t ~or t~e purpc~ses of ttle ~ e oY~?niew, o~ t~le ~er,ter fre~lency fo Of ~e ~tu~l ~701~X generated S~glia1 iS identifie~ a~ r~e~
ab~v~! tn~ flb~r rat~ of t~e fluid c~n then ~ c~lc-llated in ~ ~rlv~n~ion~l rn~nr~er, as also de~rib~ a~o~Fe.

- 17a C C T ~ 3 T H U 1 7' Cl 1 - 2~ 00 ~,~ . .

1~ow,~ ur~oC.~ r~or~ ~t.~ile~ e;~r~ tiorl, Fi~ur~ 7 illust.~tes h~w ~;igr~al C~ t.ur~s a,~ ~es~ril:,~c; m~th~matic~lly. Tl)e ~l~ri~lwi~lt.~ W
7~1 ~f si~;n~l 71 ic. m~ure~ ~t one-h~l~ oI the sl~n~ 71~ rrl~m~

--17~--O ~ T ~ '3 T H U 1 ~: ~ ? Zc~0~706~ P

,~ .
amplilude. ~he signal signature shape is represented by the s~ mbc~l Q. Q for any particular signal signature equal~ the hori~ontal center poinl (ro in Figure 7) divided by the bandwidth B~.
As mentioned ab~ve, ~e have determined by e~perimenta~ion snd analysi~ Iha~ vorte~ signal signatures have a broad bandwidth which ha~ a correspondingly lo~ value Q ~'e have found that a Q Or 2 to 3 i~ common for ~ vorle~ ~ignature in ~he frequency domain. Noise ~as rol~nd almost alway~ to haYe a spiked ~ignature with a relatively high Q Or 30. Al~o~ some noi~c ~as round to have r~ndom patterns ~ hich do not re~ult in any ~ignature ~uch as that sh~un by ~ignature 71. Nci~e wa~ al50 found to ha~e very Ic~w a~d ve~y higl~ rrequencies which shc~ up at the extremilies Or ~he horizontal ~requency a~is. The n~ise~ at Yery high and very l~w f~equencies can be removed by conventiona~ electronic high pass and low pass ~ilters.
The pre~ent in~ention uses dlg~tal signal processing technique~ to remove all other types of noi~e signals ~hal have frequencies closer lo, and ~hich ma~k more errecti~ely, the vorte~ signal.
~igu~e 8 illustrates the use oî cross-correlation to determine ir a vorte~ signal, wllich might be similar to Ihat sho~n in such as Fjgu-e 8(8), imbed~ed in a comple~ noise gignal, ~ulch a~ thal shQwn in ~iRure 8(b).
ZO 'rhe mathematical definition Or correlati~n is rt~)= lim ~ ~T
~ ,~ X(t)y~t~W)dt . (~
where r(w) ~ the correlation funclion fortned by ~ummin~ ~he lag~e~ , f ~/
product~ Or two waverorm~ ~t3 an~ ~(t~J

- 18 - . :

O 1~: T -- 1 :Z -- :3 ~ T H U 1 ~: ~ :Z 2~C)Q700 P . 2 W ~ lhe time Iqg~el~e~n Y(t) ~n~ ({ ,) ln actuality, correlation i~ a ~imilarity le~t belween ~vef~rm~ lr the ~4~ .J = Ci l~o ~verorm~ ~re th~ ~m~, ~(t) - y(l~ ihen their correlation i~ rererred to a~ an ~utocorrelation. If the two waverorms ~re diffe~ent a~ In ~igures 8(a) and 8(b!. then tt eir correlstion i~ re~errea to a~ cro~-correlation. Thus, in the present invention a narrow ~and noisê signal signature or templale 8!~ in Figure 8(a), l~ cros~-cotfelated wilh the noise signal or Figure 8(~). rhi~
resulls in ~omethin8 sin~ilar lo Figure B~c~, ~hich ~how~ UB that indeed ~any noise ~ignal ~ignatures of Figure 8~a) are imbedded in the 9ignal of Figure 8~b).
rhe proce~s Or cross-correlation multiplie~ the ~ignal or ~ig-lre 8(b) ~ith delayed versio~s or the ~ignal si~nature of Fi~ure 8~a~. The resultant croa~-correlation ~hown in Figure 8(c! contain~ only tho~e rrequency oomponent~ common lo both ~aveforms.
Il is sppropria~e t~ note at this point that the fl~ndamen~al mathematical principles utilized in the pr~sent inven~ion, including Fourier analysis and cross cor-ela~ional techniques are well-knawn and ars not de~cribed herein. ~Iso, Fourier analysis includes rasl Fourier transforms (~PT3, ~hich formulas have developed since aboul 18~2. They allow Ihe transror~ation of physically 20 ~eco~nizsble time-domain waverorms, su~h as ttle number Or ~or~ices per second in th~ application, into rreguency domain ~averor~, 5UCll a~ relaling Ihe ~mplitu~eof the vortice~ to ~he frequencies or ~heir occurrenoes. FFT formulas are ~ell kno~n in the art o~ digita~ signal prooe~iDg.
Cross correlation teçhniques afe known and used by persons skilled in the art ol digital signsl pr~ces~ing to denote Ihe similari~ies ~r t~o waver~ra~s.

- lg-O C T ' 1 ~--8 ~ T H U 1 ~: ~ ~ 2~ 700 J
Specil ically, the similal lties bel~-een t~o ~ a~efor ms can be found by su~ min~ the lagged products or each waverorm. Functionally, ~)rrelati-)n ~an ~e thoupht oJ as a n~atching up Or waveJorP~ component~ or 8 simil~rity test belween ~vaverorm~.
~ s ~sed herein, cross correlation i~ applied to determine ~hether the kno~n shape of a vorte~ freque~ ignature is hidden in the complex waverorms of a signal having both YorteX and noise co~ponenls. The rirst step mathematical~y i~to convert by meang Or FFT Ihe combined vortes and noise wave~orm rrom ~he time domain to the îrequency dom~in. ~he con~bined ~aveform i~ ~hen slored in rrequency domain patlerns. The combined waveform is then multiplied by 10 delayed version~ oî kno~n nat ro~ band~idlh noise ~ aveforms or te m plates The res~ ing cross-correlation wa~eform conlain~ only lhose rreq~lency comp~nenls common to both waverorms. 'rhus, the noise ~avefor~ component can ~e pulled rrom the jumbled hash of the combination vorte~ and noise ~aveform. One type or cro~s-correlalion ~athematics may u~e the de~ayed waverorm of the sarne ~averorm as a product. Thls product is called an autocorrela~ion. ~hen ~he ~u~ocorrelation is processed by mean~ o~ an F~r, the re~ult is cal~ed a po~er spectral density (PSD). Yarious computations using P~T, PSD and cros~-correlation methodolo~y ~re used in Ihe present invention. Numerous treatise~ are ~railab~e to describe 1he precise rO~ mulas neoe~sary t~ e~eute tho~e ~ell-kno~n techniques 20 Iha~ are applied in the present invention. The ~e~cription herein is limited to the new, u~erul and non-ob~rlou~ embodimentt Or FF~ and cro~s-cc rrelational melhodo~ e~ a~ applied to flow measurement.
Figure l l i~luslrate~ the bardware utili~ed in the present invenlion to esecute Ihe proce~ ~teps of Figures 9 and ~0. Constant cu~renl source 4.1 provides a constant alrrent Or spprogimately l.0 mA to the sIrain gauge ~ensor ~.2.

~ ~ --R ~ 1 H U 1 .-: ~ ~ P
2~70.0 The resi~tanc.e o~ Ihe strstn gauge ~ensor~ies ~ilh the derlection or the ~ing ( 59~escribed above) in response to the vorte~ slgn~a~s, and noi~e. With a con~tan~
current ~k~wing through the st~ain gaug~, a slgnal is produced, which has a vc)lta~é proporlional to its resista4n8 and a frequency indicative Or the vibr~tion~
or movements Or the win~ot sho~n in Figure l I ). Pream plifier ~.3 a mplirie~
the strain gauge signal to a le~tel acceptable to ~he low pass ~ilter. The gain of this preamr,lifier is prererably in the ran8e Or ~bout 1000, nnd the out~ut voltage le~rel can be I rn~r to 2 volt~ peak-to-peak. Low pass ~iller. 4.4 can be, for e~ample, a 4-pole low pa~s aeti~e fllter ~ith a cutorf frequency or I S00 Hz. Thi~ is the ri~S~
tO ~taBe o~ signal riltering, an~ it supptes~e~ high frequencies th~t are o~side Or Ihe l~G~ ce~
norn~al e~pected range orfre~uencies generaled by Ihe.Y~rteYe~ the ~Ibwmeter.
The output oî Ihe lo~ pa~s rilter ~1.4 is amplified by an adjustable gain amplirier f.5, which may have a gain a~ju~ted un~er rirm~a~e control to any value ~om l t~
about 1023. This gain atlju~tment i~ desirable to accommoclate a wide ~ange Or vorte~ signal amplitl~des.
The signal is then fed to ~ p~ogran~mable anti-aliasing low pass filter, 4.6, which can be .a 6-pole, s~itched capacitor lo~-pass rilter, with a cutorf fre~uency set by ~ clock input rhe f~equency of the clock signal may be 100 times Ihat Or the cutorr rrequency. rhis ~e~uency is set by the p~ogra~nmsble ~quare ~ave 20 ~enerator ~.12. rhe anti-al~asing filler clock feed-hrough fil~er 4.7 remo~es cloc~
pulse~ in the output Or the progra~able anti-~liasing ~iller 4 6. These pulses a~
at a frequency of 100 times the cutof~ frequen~y Or the anti-aliasing rilter 4.6. The signal then passe~ through a high pas~ filter ~.8, which can be a sin~le-pole RCfilter with a cutorf r~equency Or 0.~ Hl.

O C: T ~ 3 ~ T H U 1 ~: ~ 5 Z~Q~70~ ~ . 2 3 ~ he le~el ~ er 4.9 take~ a bipolar signsl and c~)nverls il lo a unipcJlal~
~ignal ror prc~es~ing ~y the A/D converter ~.14. The inr~t l~ange or the le~el shil~er u~e~ in this embodiment is -2 to ~2 volts, ~ith an output r~nge o~ O to 4 volt~
~ peak de~ector 4 10 is used to sense the leYel of the smplificd signal ~efore it is filtered by the anti-ali~sin8 filter 4.7. r~his step is desirable to prevent saturalion Or the anti-aliasing filU~4.l~? The pé~k detector oulpl~t is sampled, and is u~ed for ~etting the gain Or Ihe adjustab~e ~ain ~mpliJier 4.5 under rirm~are control IQ The analog switch ~ used to select, under contro~ of the firm~are,eithel~ the filtered sign~l roa ~ampling, or the pe~k ~etect~r ~t~ut. A
p~o~rammable squa~e wa~re gene~at~l 4.12 pr~vicle~ a sclua~e oulput ~ith a tre~uency prererably between 10 kHz and 320 ~H~, controllsble ~y the rirmware.
rhi~ îrequency is ~Ise~ di~ecl~y to control t~e cutorf frequency of the anli-aliasing IOW p8SS filter 4.7, and it is fufther ~ivided by the ~ample and inters~upt generator {.13. rhc ou~put ke~uency of the programmab~e squ~re ~rave ~enerator 4.12 is divided by 40 lo provide timing rOr ~he analog to digital converter ~ during ~ignal samplin~. Tt~is genera~es a simultaneous in~errup~ to the T~IS320C10 digitai signsl proces~or 4.18. The 10-bit ~nalog to digital con~ erte~, ~t a rate determined 2û by the programmable square ws~e generator ~.12, oonverts the amplifled and riltered ~nalo8 ~ignal to digi-al rorm~ ~hich i~ then read by the TMS320Ci O ~igital ~ignsl procc~s~- ~.18 A user progr~mmable input oomprises a set Or range switche~, which Inay be progrà~m~d by the u~er to ~elect the maYimum range, the range oî fluid density, ~nd ca!ib~ation coe~icents Or the de~ice. A ~ead only program memory comprises O C T -- 1 ~--~ ~ T H U 1 ~ ~ 2 3 ~ Z~700 8192 16-bi1 ~or~s Or ~ead only memory containin~ the prograr~ inslruction~ for the TM~320C10 digital signal processor 4.18 These ins~ruclionS ale also rererred to as firmware. A read/write data n~emory comprises 81~ 16-bit words o~
~ead/write d~ta n-emory and is used for storage Or the large sampted dat~ and spectrum arrays. The digital signal p~ ssor ~an be a TMS320C10 digit~l si~nal proeessing microp~ocessor integrated cireuit. The control logic 4.19 provides the following r~nctions-a. Input/Oulput por~ strobe circuit.
b. Anti-aliasing filter clock reedthrough rilter data latch.
c. Analog switch select la~.ch.
ro~ran~mable square ~aYe generator eour~ler l~ch.
e. P~ogram men~o~y page selecl la~ch.
f. Rea~/write data nleo~o-y ad~res~ ch.
8 Outpl~t circuit type sen~e.
h. Output circuit ~at.~ Iatch.
t. Watchd~g timel circuit.
There are t~o types of output circuits ~.20 -- sQuare waYe ~nd analog. The unit aut~matically ~en~es which circuitry is ins~alted, ~nd pro~ides outp~t signals p~oporlional t~ the Mow.
~erore referring to Figures g and 10, a rurther o~erview and definition of term~ is pro~ided.
Digital Signal Pro~essing - The c~noept~ Or dig~al signal processing h~ve been e~lablished in ths last ~or~y years and are ~ell documented in te~tbooks.
~ome Or the term~ used here ate men~oned to xi~e a brief terPlinol~gy reference.

O C T -- 1 ~-- 3 ~ T H U 1 ~: ~ ~ 2~Q700 P . 3 ~3 ~ `~/
l~nlt.~i~ used to clent,le Ihe dimensionless measure Or the f~e~uency ~c main The digilal signal processin~ lechnique is based on sampling at a selecta~le s~mpling rate. ~erformin~ a 102~-point comple~ P~ resultt in 1024 CO~ple~ arra~
elements, Or whieh only ~he first 512 are needed lo creale a real po~er spectrun-.
These 512 are described as having their "horizontal a~is" in diménsionless u~its. 1~
unit~ correspond~ to an actunl lrequency at the ~ampling rate (in Hertz) multiplied by ~/I()2~) wh~re N 1~ be~w~en 0 and 51 1 1 rhc ~Ignal r)r~e~lng l~hnique~
u~ed here are independent Or sclual sampling îrequency. so ~11 lhC rcrcrencc~ ~re kepl as dimensionless units.
G~o.~ t7rreht~d,~? is ~ mathemalic~l proces~ ~here the inpul is two real array~ and the ~utput is a single resl array. The output ~ends to be larger where both inpUI array~ have the same shape. ~ "direct cr~ss~ rrelation" is used on snlall ~rray~ because it is the raslest method for them, and ~u~es all real arith~etic Pa~t convolution" is raster and thus used in performing a cross-~rrelati~n on large arrays. The compleY FFT o~ each real array is ~oa~puted, the l~o arra~s are comple~-mulliplied, and Ihen an in~er~e-~FT of the c~ple~ resull array is performed to Bet the cross-correlation array tesult.
~ is a tneasure Or lhe 3-dB ~andw}dth of a signal, i.e. the ratio of the center rrequency to the dist~nce bet~een the hall-power f~eq~enc~e~ (the bandwin~h) 2n Rerer~ing ne~t ~o Figure 9 the proc~ss steps l~tiIized in executing the bes~
mode ot the present invention ~re de~c~ibed.
rhe Start 1.0 indicates the start Or e~ecution o~ the program when the unit is rirst ~ ned on Initiali~ation I . I ~omprises the fo~lo~ring steps-~. Set the gain Or the ~djustable Gain Amplilie~ 4.S to minimum.
b. Set vsl-~et ol constant~ in data me~ory ~ ~-- T ~ 1 ~ T H U 1 ~: ~ ~ 2C~ 700 P ~ 1 ~ ~ .,i c Read the u~r input 4 ~5. ~hich determines rna~imum fl~ ate, ~en~ity range at~d ealibration c~erficients of th~ unit;
d ~et the Programmable Square Wave Generstor(~f ~equency based upon the maxirnum flow rste.
e. Set the Anti-aliasing Filter Cloek Feedthrouxh Pilter ~ 7 rrequency;
r Determine the number Or power ~pectra to aversge during calculation cycle. ~his num~er wi~l ~ar,ge rrom 10 to 18, depeding upon the ma~imum flo~ rate programmed ~y the u~er.
In the Initialize Ca!culation Cycle, the r~llouing initialization is done:
a. Sel the Analog S~itch 4.1 I to select lhe Peak Detector 4 10 o~tput;
b. Set ~he guin Or the Adju~table Galn Amplifier 4 ~fased upon ~ke p~ak detecto~ output;
c. Se~naJo~ S~itch 4.1~ to ~elect the ril~ered ~ignql rrom the Level Sh~ler 4.g utput; ahd d. Initi~lize all data arrays and set the number ~f spectra ac¢u~ulated to ~ero.
In~tep }.3, Sa~ple 1024 Point~, a~ a rate determined ~y the OUtpUt frequency o~ the ~ogrammable Square Wa~e Generator ~.12 and the S~n~ple and Inlerrupt Gener~tor ~.13, the filtered ~Inal~ signal i~ oonverted to digital dat~ ~d 20 ~tored in the samp~e~ data array.
The po~er ~pectru~ i9 calculat~d at step l.~ ~y doing ~ ~a~l F~l~rie~
transform (FPT) Or lhe sampled data, and then suP ming the sql)a~es Or the ~eal and imaginary components. Since l~e ~ampled ~ta i~ real, the re~ultant po~er ~pect~um is symmctricsl, and Ihe ~elative po~er at r~equencies fron~ 0 to 51 1 units are u~ed.

O C T ~ ~ T H U 1 ~ c~ ~ p ~ ~
Z~ 700 The p~er speCl~Um i~ then a~de~ into a sum or p~ue~ speetra arrsy in ste~
1 5 for 1a~er averaglng. By averaging several spectra, Ihe signal-to-noi~e ra~io jc impro~ed, and spectral jitter is reduced. A typical unaveraged power spectrum i~~ho~n in F-gure 1~.
Rererring aBain to Figu~e 9, the number of ~pectra accu~ulated is co~pare~
at ~ep 1.6 with the nu~nber det~rmined during initialization ~lep 1.1, and~ if no~
equal. continlle~ with step~ o 1.6.
In step 1.7, Calc~lale Aversge Power Spectrum., the spectral values i~ Ihe sum Or po~ter ~pectra array are divided by the number Or ~pectra accumulate~, 10 resu1ting in lhe average pawer ~pectrum array a~ illustrated in Fi~ure 13 AJler noi~e riltering in s~eps ~ 1 lO 2.11 and signal es~im~tion ~teps 3.1 to 3.12 inFigure lO,thereslllt,~epresentingtherlowr~ ou~put at 1.8 in ~igure 9 either as a ~qusre w~ve with a rrequency proportional lo the rlow, or as a 4-20 analog signal proportional to ~he rî~
Bl~ck~ 2.1 through 2.l I represenl the noise rillering s1eps. In sun~mary, lhe noise removal ~ steps 2.1 lhrough 2.11 is ba~ed on si~nal templates ~ith a Q ot 3, ~nd noi~e ~ei~plates uith a Q of 30. rhe progr~m rirst removes the lo~ frequen~ynoi~e trom the spectrum. rhen it perrorm~ a cro~s-correlation betueen a noise lemplate and the spectru~n. Thi~ help~ lO i~et~tify location~ of possible noi~e peaks 20 ~hile al~o per~orming ~ome spectral smoothing. Peakt in the ~ross correlationres~llt represent possi~le noi~e peaks. These peaks are evaluated ~y checking whelher lhey more closely match the signal lemplate or the noise lemplate. If they appeal~ to be noi~e peaks, Ihey ~re remo~ed ~nd rep~aoed by a linearly interpolated set of' points. This program perrora~s this noi~e removal algorithm O C T -- 1 2 -- ~ ~ T H U 1 ~ ~ p, ~ 3 Z~ 70~

l~ice. ~irecl cross-correlstion i~ u~ed ~or all noi~ ~illerin~ Junclion crn~-cor~elations.
Sp~cifically, in ~tep 2.1, Remo~e Low ~requen~y Data, the criteria for removal Or very lou frequency not~e i~ lo ehec)~, beginning al a rrequency Or ~ unit.
Io ~ee whether the ne~t value is smaller than the current value. ~'hen the ne~t vallle is not smaller th~n the curren~ Y~lue, ~he checking process stops. The current vall~e becomes the top endpoint r~r interpol~tion. The bottom endpoint ror interpolation al~ays has 8 val~le of 0 at 0 unt~s. Any points between 0 ~nd the 10 endpoinl ~re linearly Interpolated (straighl-line interpolationJ.
The Noise Te~plate Cr~s Correla~ion s~ep 2.2 involve~ cross-~rrel~t.ing ~
~eri~s of prererably si~ noise templales ~e lasl five Or ~hich have a nominal Q oJ
30 an~ are re~peclively oentere~ ~t 20, 40, 80, 160 and 320 ~nits) wi~h the ~ianal spectru~. In this case, the result i~ called the noise cross-corre}ation array. An e~ample o~ lhe re~ult i~ illustraled in ~igu~e 14.
'rhe ~ros8-corlelation is comp-lted in ~iY pieces, one piece rOI e~ch noi~e template. rhe start and end points for each piece are: template ~1, O to 13; #2, 3 ~
to ~7; #3, 2B to 56; ~, 57 to ~12; ~5, 113 to 225; #6, 22~ to S10 units. For template~ ~2 Ihru #6, their start and end points ~ere determined as follow~s. The start points are each template'~ center divided by the ~quue root Or 2, and the end 20 point-s are its oent~r times the squ~re root Or Z, ~ith the special c~e Iha~ the te~plate oentered at 320 h~ an end point at 510 tJnit~.
The template values ~re ~tored con~tant~, stored already normalized, pacl~ed togelher into ~djacent ~rray elcments. rhe te~plales are ch~en ~ith 8 size e~timate ba~ed on the I O peroent power le~el Or the function. The function used to oompute the value~ Or the templa~e i~

O C T ~ 8 ~ T H U 1 ~ ~
~ 70~
.- ! (33 )oQ~ A' where 8 Is the normali~ed frequency ~a=f~o, ~here fo Is Ihe center kequency), and Q is 30 ror nol~e templales. The templates are meant to be used Wlt1~ directcross-correlalion. rhe tetnplates are each scaled to unit area ~y ~;eeping 8 running sum o~ the values and then dlvldlng each Value by ~he to~al sum. A temp~e ~tr~ctl~n method is used that estrscls lhem directly out, one at a time when neeaed.
1~remplate ~1 ls a single poin~ templale, s~ ror slmPlicity the p~wer spectrum element~ O thru 13 a~e copie~ int~ Ihe cross-correlation ~esul~ array.
~or the l~st rlve templates, the piece or Ihe cross-correl~lion that is perrormed is dPne In t~e rollo~ing manne~ Glven a template consisting Or elements t~il, and a power spectrum densily c~l~slstlng or elements psdlil, the direct cross-co~relatlon array consisting o~ elements cltl is computed rOr each Yalue or t tlhe ran8e o~ t wlll ~e the ssme as the r~nge Or the template tp) as rollo~s:
~ sua~mation is perror~ed o~er the rsn~e Or i rrom su~mation st~rt to o,ummation ena. Summat~on start is at l-~enter, and summation en~ Is ~t end~t-cen~er. rhe summation is or t~e produ~ts Or lpli-l~centerl times psdlil.
~0rhis pro~e~ure ~ e~ecuted twice, and an eYample ot Ihe resul~ or the sec~nd pass is lllustra~ea in ~igure ~5.
Rererrlng a8ain to Figure 9, In the step Z.3 labeled Find Minimum sn~
Maslmum Values, all reI~ti~e mlnlmum a~d ma~lmum values In ~he nQise ~oss c~frelation al~r~y are found. The rre~ueneies, ln units, at whlc~ the mlnima andtna~ima oocur are stored in the ~ ma~ array. Ne~t, ~ step 2.4, ~C ~ ~requency at - 28 -O C T ~ ~ T H IJ 1 7: 5 ~ 1~ . 3 5 ~, 2~ 700 ~_J
Firs~ Ma~l~um., Ihe proce~s o~ ~in~ing and removin~ noise pcak~ is per[o~med al the ~ocation or each ~a2in~um, ~tarting with the IQcation of the rir~t maximum value ~oun~ in the noi~e cross-correlation a~ray.
In ~etermine Valid Range, step 2.5, the valid range e~tends 8S wide a~ it can rrom the lar~er of:
1. The location Or the mlnimum al the left side Or th~s ma~imum Z. rhe location of the lert ~ide IOx levei o~ a "n~ise template" ~ template with a Q characleri~lic Or noise ~c~ Qn~ 30 ha~ been used h~e b~l other value~ could be used; the lempl~e is centered at ~he current ma~imum, and has a 10 ~lue Or unity at it~ center);
o Ihe smaller ol-1. rhe location of the ~inimum at the right side Or thi~ ~2alimum, 2. The location of ~he rlght side IOX le~rel of the "noise template' centered althe current maIimum, 3. 5 t I units rhe location Or the lerl ~ide I OX level Or a "noise template can be accur~tely e~tionated for Qn greale~ Ihan 2 a~ the truncated integer value o~ center rreQuency, divided by Qn. times ( I minu~ a factbt ), br;

(4) 1~1~;~7=c~ll ~

The ractor is dependent on the Q ch~ract~rlstic o~ noi~e In the ~ollo~lng ~anner:
(53 - 2g -O C T -- 1 2-- 8 ~ T H U 1 7: 5 1 p, ~700 ~hcr~ P ~tands rOr the p~wer level Sinc~ P=O.I (derived from ten ~ercenl Or unity ~hen the templ~te hs~ a value Or unity at it~ center), then if Qn - 30, the raclor i~
0.05.
Si~ilarly the loeation of Ihe right ~ide J Ox level of ~ n~ise template can be accRrately e~timated ror Qr~ Rreater than 2 a~ the ~runcated inleger value Or center frequency, dividé~ by Q~, times ( I plus ~ ractvr~, plu~ one, or:
~6 ~he nesl ~lep 2.6 is to cniculate ~pectrum fi~ to templates. OYer the Yalid ran8e the shape o~ the aclual dala is evaluated The signal is assumed t~ h~e ~
nominal Q of 3 ~hereas the noi~e i9 assumed to have a nominal Q o~ 30 as specified abo~e. Ir the noi~e model proYides the ~elter match to the data in the valid range, then po~nts ~ithin the ~al~d range are marked rOr r~oval.
If ~he valid range has only 3 point~, interpolation ~ a nnore accurate estimate Or the location Or the peak ba~ed on a be~t ril Or ~ noise template to the data i~ perrorméd. This step is n~c~ary for rea~onable discrimination between signal~ and noi~e when the noi~e peak lie~ be~een adjacenl spectr31 poinl~. The 20 peak i~ inte~pola~ed to an acwracy Or 1/~ unit, by co~paring its fil to lemplates between lert and center at 1/4 unit interval~, and then between right ana center.
rhe fiI Or the po~er Spectrum to the noise template is t~en compared a~ Ihe s~ep 2.7, Be~ter Pit to Noise Tetnplate?, ~o the ril Or she power ~peclrum Io the 5ignal ~emplate. Thi~ i~ 8 dlrect co~pari~on of two nu~eric values. Ir the ri~ 1O the noise templ~te i9 the larger Or the two nu~bers, shen the ~e~step i~ perrormed, O C T ~ ~ T H U 1 ~: 5 :Z 26~30700 ` rl - ie, Mflrk Points ror ~e~oval 2.8. In thi~ neY~ slep 2.8, il the ~It Or ~he p~er ect~um to lhe noi~e template i~ higher th~n its fit to 1he signal lempl~te, ~11 point~ in ~he valid range (determlned in step ~.5) ~re marked ~or re~oval In rd~iti()n, if lhe tert mlnimum point is surrounded ~y points marked for removal, it should al~o be ~arked for re~oval. Steps 2.5 through 2.10 are repeated for all m~imur~ values in ~he min-ma~ arl~ay, a~ the FC - ~requency o~ Nell h5a~imum 2.10 is incremenled for e~ch cycle.
t~hen the proces~ i~ done ~hecking all maYima, as ~h~wn 8t 2.g, il then proceed~ to Remove Ma~ked Point~ 2.11. ~he location~ o~ all or the pOi~lS lo be 10 removed are stored in the remove array. ~or each seql~enCe Or points marked to be remove~ the points in the power speetrum are replaced by ~ straighl line. 1~ thema~nitude of a point to be removed 1~ than the interpolated str~ight line Yalue~il i~ not replaeed.
Re~erring no~ to ~igure 10, Blocks ~.1 thr~ugh 3.12 r~presen~ the signal estimation steps. This rouline performs signat estlma-ion for the insertion vorle~
meter. The averaged, noise-~iltered power spectrum densily is cross-correlaled ~ith seven ~ignal templa~es rO~ a rirst-pa~s estimste. A signal tempJa~e is thenconstructed ~ that frequen~y and i~ agsin ~o~ orrelated wilh he spectrum f~
up to two iteration~. The result is then ~urve-îitted to determine the si~nal peak 20 with better re~olution. A ~inal ch~cl~ Is made by verif~yinB Ihat the result is cQn~istent ~ith Cl'iteri8 0~ signal ~trength an~ noise ~i~imizalion.
In Signal remplale Cros~ Co~reJation 3.1, Ihe ~i~st pa~ templste method requires e~traction o~ ~even templates to c~er the rull spect~um. ~or e~se Or ir~plen~entation, the templates are equa}}y spaced on ~ log scale by a r~ct~r o~ 2.
The teahplate~ have cente~ frequencies o~ S, 10, 20 ,40 ,80, 160 and 320. The -- 3~ --n c T ~ 8 ~ T H I I 1 ~: 5 :~
2~70~
~. ,~
template~ are no~malized to unity area ~sum) so tha~ they arc nol biased ~hen u~ed with ~ ~vhite nois~ b~ckground. A valid ran8e has be~n established rO~ eachtempla~e ~rom ~tO
Each Or the te~p!ates is partially cros~-correlate~ with the ~pectrum. This i~
done directly bec~use it is erricien~ ~or the lo~ rrequency t~emplate~ ot 5, 10, ~0.
40, ~nd 80 units, and via rsst convolution using the FPT, whlch is n~ore erricient for he higher f~equency 160 ~nd 320 unit templs~es. 'rhe working template ~a~
g~ne~ated by taking the sign~l te~nplate and circula~ly ~hif~ing it so that its peak i~
at zero. When the convoluti~n is employed, mirrored v~rsions Or templa~es are 10 u~ed. Mirroring Or the template ~efe~ to mirroring a~out 1he c)risin.
Wil~in the valid range fo~ each template, the largest valIle in ~he cro~s-corr~l~tion ~n~ the corresponding îre~uency is checked IO see il lhe value i~ aslflrge 8s the previous n~a~i~um. Ir it is, then the frequency is assum~d to be located close tO ~he actllal signal flnd the frequency corresponding to this l~rg~st value i~ chosen ~s the first ~ignal ~tequency estimate. In the e~tremely u~likely condition that mul~iple frequencies end up ~ith the same value the lowest frequency Or the IWo i5 a~sumed n~ost accurate.
~ he inputs ~o this routine are the po~er spectrum den~ity and the comple~
FFr Or it. 'rhe comple~ array i~ e~pecled t~ tave the rormat required ~y the 32010 ZO assembly langu~ge FF'r rouline, whcre th~ real snd imagin~ry par~ al~e in~erlesved. 'rhe power spectrum den~ity i~ intende~ to have only 512 points used o~Jt of ii; the other 512 are e~pected to be zero-~illed. The rou~ine uses the ROM
templale. rhe in~e~er f~rst-pass signal ~requency es~imate i8 returned. This rou~ine al~o produces ~n ~ver~ge cro~s-corre~alion a~et age Or the cros~-correlation over Ihe entire ~pectrulP range it e~amines, for u~e in the rinal check routine 3 ~ 0.

O C T -- 1 -2--~ ~ T H U 1 ~: 5 3 2~0C~706~

, ~ ~, ln Set IT~F~AII~ tep 32, .Ihe pr~cess oî rinding the signal peak i~
~errorme~ twi~e if needed, and an indicatol rlag is needed. Here it is ~el to indica~e Iha~ thi~ is the first iteration.
1~ n Set C~N~ER = ~re~uency ~t Ma~lmum Value of C~o~ Co~relation., slep 3.3,~
the purpose is to rind the ma~imum val~Je in the cross correlation arr~y, ~n~ set ~NTER equal to the fr~quency al which this maYimum ~alue occurs. 'rhen In step 3.~, Cross Cor~elate Spec1rum ~ith Signal remplate al C~NTER, once a ~irst-pass signal frequ~ncy estimate 'peak'` has bee~ msde, a ne~ signal template is genera1ed at lhis rrequency. This ne~ template i~ then cro~s-correlated ~ith theiO specl~um. A mirrored, ~ero-eentered lemplate is u~ed in a f~t~ ~onvolution impl~m~nt~tion of the cross-obrrela~ion in order to end up ~ith ~ peak near th~
~ignal tea~plate peak. ~l~e ~rr~y `~" ha~ ~Ire~dy been FPT'd, so it iS ready ~orconvolution with the lemplate ~hat gets ~enerated.
In ~tep 3.S, Is Cross Correlation Peak Close ~nou8h to Center?, the ~onvo~u~tion resull is checked to find ~he peak in a 2-i range around the signallemplate peak, rro~ template pea~: divided by the square root Or two to templ~tepeak times the s~ua~e root of two. Ir ~he convol~ed peak is wilhin '`one peroent or one uni~ (whichever js grealet)" ~ Ihe sign~l template peak, then the ne~t step i~
inlerpoîation lo the sctu~ peQk. If it i~ not within those litnits ar~er one iterstion, 2Q then the convolution is repeated with a new template at the new e~tima~ed signal emplate peak. If ~ter the second iteration it i~ ~till nol ~ithin hose limit~, then hiS i~ a r~ilure. rhis rouline checks whether bpeak is close eno~gh to apeak. rhe apeak i~ idered to be a ~i~nal templste peak, whi~e bpeak is the ~e~ul~
convolution The bpçak should be apeak ~/- I unit, or bpeak ~hould be wi~hin onç

~ , Z~ 700 percent or ~peak, whichever is lar~er Ir bpeak is close enbu~h the routine returns a l; i~ not clo~e enough, this rouline returns a 0.
Step 3.6, IrER~lON = 1?, is pro~ided so that the pr~ess or finding the signal peak is perr~ormed Iwioe, if needed, ~nd an indicator rlag is neecled Here i~
is checked to determine if it i8 the rirst ite~ation. Step 3.7. Set l~ERA~SON ~ s provided ~o that Ihe process Or finding lhe signal pe~k is pe~formed t~A~ice, itneeded, and an indicator flag is needed. }lere it i~ set to indicate tllal ~his i~ the ~econ~ i~e~alion.
ln step 3.8, ls Spectrum ~ata Satisfac~bry for lnlerpolstion? first, a check is 10 made to rind the peaks in the convolution result. ~f two adjacent peaks ~e found, then the inlerpolated peak i~ ~t the eenter of them. Ir more than t~o adjac.enl points are equa~, or t~o non-adjacent peak points are equat, then this is a failure, and the outplJt is ~et as iî no flow (zero rre~Jency) occurred. Also, ir the tailure indicator ~s set ( second peak w~s Q) Ihen lnte~polalion is not perrormed.
Generally, in step 3 g, interpolation is performed by 8 parabolic curve rit ~o ~he three point~ nearest and including the peak ~peak . Ir only one point is thepeak (~hich will almost alway~ be the c~se) then a parabolic curve fi~ is perro~med u~ing Ihree point~, Ihis polnt ~nd the polnt on ~ach side. This parabol~c curve ril uses the derivative~ from Crsmer~ rule for solving the three ~imultaneous 20 equation~ o~ the fo~n~
y-a~2~b~c (6) Witl~ the solutio~ ~ = ~ (7) 2- Dl wbe~e D~ x~ (x~ 8 ) F T -- 1 2 ~ ~ T H U 1 ~ !5 cl. 2~0~7C)O
~ , , snd ~2=X~ U~ ) (g) The solulion ~o i~ the frequency peak bet~7een ~he p~inls (~I-YI). (~2.Y2). and (~3 Y3) The final check at ~tep 3.10 determines ~hether th~ ~esul~ Or the entire îunctional algorithn~ is rea~onable.
In the steps Second-Pa~s E~timate and Interpolation to Si~nal Peak there ~ere fa~lure conditions lr alter t~o steps Or the Second-Pas~ E~timale the result is still not within reasonable Jimit~, then this was a failure lf there are the wrong 10 lype of peaks in Ihe Interpolation to Signal Peaks, then IhiS ~as a railure. Ir eilher of th~e condition~ oceur, the output wa~ ~et as if no r~O~ (zero rrequeney) occurred. ~ho~e checks were done in tho~e roul-ine~ and are not done in thi~
routine. / "~ 5 f~S~
rhe peak "fre~eak" from the Second Pass step m~t be a~ least a r~e~uen~
dependent ralio time~ the cross-correlation aversge (rrom Pirsl Pa~, o~er the entire range) value of the cro~s-~orrelalion ~unction, in o~der to pre~lent very sma~
cros~-correlation peaks from being incorrectly recognized as signals. Note that ~his 1~ nol the interpolatèd peak value, onîy becau~e ror it eYtra computation o~
derivalive~ would have been ne~essa~y in the interpolation ~tep, and t~e 2econd ~û pa~ peak value i~ e~pected to be very cloge to the (not-calculated) interpolated peak vall~e. Iî the peak i~ les~ than the r~tio tin~es the average, then lhe output is set 8s if no flow (ze~o rreqllency) occurred.
'rhe final rejection c~iteria in step 3.10 compri~es l~o chec~s. rhe rirst is 1lckeck of the signal amplitu~e ~s a ~unction Oî density ~nd ~requency. The ~econd is a check d the the signal rrequency a~ a runclion Or density - 35 - .

O C T -- 1 2 --~ ~ T H U 1 ~: 5 5 2~?0Q700 ~ P ~ 1 V ~
-Firs1 Chec~
The ~8na~ power musl be ~ea~er than a minimum limit ~hich i~ dete~mined a~ a runclinn of densi~y and frequency. The magnitude Or the po~er spcctrum at the e~tim~ed signal rrequency i~ compared with a threshold value ~hich is determined from ~he t~uare Or the densily sn~ lhe ~ourth po~er Or the estimated rrequency. Ir the magnitude i~ le~s than the thre~hol~ value, the OU1pU~ is sel a~ ir no flow (zero frequency) ~ccurred.
~ . Second Check The signal rreqllency must be greate~ than a minimum limit whicl~ is a lable 10 Inokup determined a~ a fun~ti~n or density, ~l~e density i~ sho~n in pcJl1nds per cubic root.
~n~ity Mini~um.
I~/rt3 Frequency, Hz S,o.l 150 S 0.25 7S
~ 0.5 ~0 Ga~,>0.5 38 Llquid 5 1n ~tep 3.11, ~e~ult ~ Interp~late~ Peak, if the rinal check criteria a~e pa~sed, the l~e~ultanI frequen~y i9 as~umed vali~ for outp~ltting. Ho~rever, ir the rinal check criteria sre nol pas~ed the output i~ set as ir no flow (zero rrequency) occ~rred, a~ si~o~n Rt ~.12.
rhe last sIep Or the proces~ simpiy calculates the rlow rate rrom Formula (t), i.e, Fl~ K Frequency in Ihe Mgit~l ~ignal Proce~or ~.18 of O C T -- 1 2-- ~: 5' T H U 1 ~: 5 ~ p, ~ ~
26i ~3~700 ~ , ~l~ure 11 and OUtpl)tS ~ u~eable signal by means Or the oulpUl circuit ~ 20 c)f Figure l l The roregoing is con~idered as i!lu~t-rative only Or thê principles Or Ihe inven~ion. Purther, since numerou~ modification~ and chan~e~ will readily occur to those skil!ed in the art, i~ i~ not desjred to limit the invenlion to the e~aet con~1ruction and operation shown and described, ~nd scGordingly all suitable modiricstions and equivalent~ may be resorted to ralling within the scope or ~he invention a~ derined by the clalms ~hich rollo~.

Claims (27)

1. A method of measuring the flow of a fluid in a pipeline, wherein said pipeline has noise, comprising the steps of:
a) means for inserting a flow barrier in the pipeline, thereby forming a wake of vortices;
b) means for providing a means for creating an electronic signal for each vortex, wherein said means for creating an electronic signal also produces electronic signals for the noise;
c) means for transmitting the electronic signals of the vortices and noise to an electronic processor;
d) means for converting the electronic signals of the vortices and noise vibrations from the time domain to the frequency domain;
e) means for removing the electronic noise signals from the electronic vortex signals by means of cross-correlational computations, thereby computing the vortex frequency; and f) means for computing the flow from the computed vortex frequency.

2. The method of measuring the flow of a fluid in a pipeline of claim one, further comprising the steps of:

g) inserting the flow barrier of step (a) further comprising a bluff body.
h) providing a means for creating an electronic signal for each vortex of step (b) further comprising a wing and strain gauge assembly;
i) providing the electronic processor of step (c) with amplifiers, filters, an analog to digital converter, and a central processing unit.

3. The method of measuring the flow of a fluid in a pipeline of claim one, further comprising the steps of:
j) filtering out any high frequency content of said transmitted electronic signals of step (c).

4. The method of measuring the flow of a fluid in a pipeline of claim one, further comprising the steps of:
k) converting the signals of step (c) to digital signals by repetitively sampling the signals 1024 times each four seconds;

5. The method of measuring the flow of a fluid in a pipeline of claim one, further comprising the steps of:
1) converting the electronic signals of step (d) using fast Fourier transform (FFT) calculations.

6. The method of measuring the flow of a fluid in a pipeline of claim one, further comprising the steps of:
m) storing the results of step (d);

n) averaging said stored results, thereby computing the average power spectrum density (PSD) of the electronic signals of the vortices and noise vibrations.

7. The method of measuring the flow of a fluid in a pipeline of claim six, further comprising the steps of:
o) removing very low frequencies from the average PSD.

8. The method of measuring the flow of a fluid in a pipeline of claim seven, further comprising the steps of:
p) cross-correlating the average PSD with approximately six narrow bandwidth (high Q) signals representative of the noise signals;
q) storing the resultant cross-correlations as noise template cross-correlations having peaks;
r) comparing each noise template cross-correlation peak first with a wide bandwidth (low Q) vortex signal and then with a narrow bandwidth (high Q) noise signal;
s) eliminating the noise template cross-correlations resulting from (r) which are more similar to narrow bandwidth (high Q) noise signals;
t) storing the results of (s) as the filtered PSD, wherein the filtered PSD has at least one peak;
u) estimating the vortex frequency from the filtered PSD by means of cross-correlating the peak(s) of the filtered PSD with at least six broad bandwidth (low Q) vortex signal templates, repeating this step, and computing the best estimate of the vortex frequency therefrom:
v) comparing the best estimate of the vortex frequency in step (u) with a minimum threshold value of the amplitude of said best estimated vortex frequency;
w) repeating steps (p) through (v) if said amplitude of said best estimated vortex frequency is less than said minimum threshold value;
x) computing the flow of fluid in the pipeline from the best estimated vortex frequency if the amplitude of said best estimated vortex frequency is above said minimum threshold value;
y) displaying the flow computed in step (x);
and z) repeating steps (m) through (y) every ten seconds.

9. A method of measuring flow using a vortex shedding flowmeter, comprising the steps of:
aa) mounting the vortex shedding flowmeter in a pipeline;
bb) creating electronic signals from the vortex shedding flowmeter proportional to the quantity of vortices produced by said vortex shedding flowmeter;

cc) transmitting said electronic signals and any other noise signals to an electronic processor;
dd) converting the transmitted electronic signals of step (cc) from the time domain to the frequency domain by means of Fourier analysis;
ee) removing the electronic noise signals from the electronic vortex signals by means of cross-correlational computations, thereby computing the vortex frequency; and ff) computing the flow from the vortex frequency computed in step (ee).

10. A method of measuring flow, comprising the steps of:
aaa) mounting a vortex shedding flowmeter in a pipeline, wherein said vortex shedding flowmeter in a pipeline, wherein said vortex shedding flowmeter further comprises a transducer producing electronic signals proportional to the shedding vortices and pipeline noise;
bbb) inputting said transducer signals into an electronic processor;
ccc) converting said transducer signals from the time domain to the frequency domain;
ddd) removing all but the shedding vortex signals in step (ccc) by means of cross-correlations;
eee) computing the flow from the remaining shedding vortex signal of step (ddd).

11. A method of producing an electronic signal proportional to flow, comprising the steps of:
i) mounting a vortex shedding flowmeter in a pipeline, wherein said vortex shedding flowmeter further comprises a transducer producing electronic signals proportional to the shedding vortices and noise;
ii) converting said transducer signals from the time domain to the frequency domain; and iii) removing the noise signals from the vortex signals by means of signal signature comparisons.

12. The method of producing an electronic signal proportional to flow of claim 11, further comprising the steps of:
iv) averaging and digitizing the frequency domain signals of step (ii);

13. The method of producing an electronic signal proportional to flow of claim 11, further comprising the steps of:
v) comparing the repetitive low Q signal signature of the vortex signal relative to the random and/or high Q signals of noise as the means of signal signature comparisons of step (iii).

14. The method of producing an electronic signal proportional to flow of claim 11, further comprising the steps of:
vi) using Fourier analysis in step (ii) to convert said transducer signals from the time domain to the frequency domain.

15. The method of producing an electronic signal proportional to flow of claim 11, further comprising the steps of:
vii) using cross-correlational computations in step (iii) for removing the noise signals from the vortex signals.

16. An apparatus for measuring the flow of a fluid in a pipeline, wherein said pipeline has noise, comprising:
a) means for inserting a flow barrier in the pipeline,thereby forming a wake of vortices;
b) means for providing a means for creating an electronic signal for each vortex, wherein said means for creating an electronic signal also produces electronic signals for the noise;
c) means for transmitting the electronic signals of the vortices and noise to an electronic processor;
d) means for converting the electronic signals of the vortices and noise vibrations from the time domain to the frequency domain;
e) means for removing the electronic noise signals from the electronic vortex signals by means of cross-correlational computations, thereby computing the vortex frequency; and f) means for computing the flow from the computed vortex frequency.

g) filtering out any high frequency content of said transmitted electronic signals of step (c).
h) converting the signals of step (c) to digital signals by repetitively sampling the signals 1024 times each four seconds;
i) converting the electronic signals of step (d) using fast Fourier transform (FFT) calculations.
j) storing the results of step (d) k) averaging said stored results, thereby computing the average power spectrum density (PSD) of the electronic signals of the vortices and noise vibrations.
l) removing very low frequencies from the average PSD.
m) cross-correlating the average PSD with approximately six narrow bandwidth (high Q) signals representative of the noise signals;
n) storing the resultant cross-correlations as noise template cross-correlations having peaks.
o) comparing each noise template cross-correlation peak first with a wide bandwidth (low Q) vortex signal and then with a narrow bandwidth (high Q) noise signal;

17. An apparatus for measuring flow using a vortex shedding flowmeter, comprising:

aa) means for mounting the vortex shedding flowmeter in a pipeline;
bb) means for creating electronic signals from the vortex shedding flowmeter proportional to the quantity of vortices produced by said vortex shedding flowmeter;
cc) means for transmitting said electronic signals and any other noise signals to an electronic processor;
dd) means for converting the transmitted electronic signals of (cc) from the time domain to the frequency domain by means of Fourier analysis;
ee) means for removing the electronic noise signals from the electronic vortex signals by means of cross-correlational computations, thereby computing the vortex frequency; and ff) means for computing the flow from the vortex frequency computed in (ee).

18. An apparatus for measuring flow, comprising:
aaa) means for mounting a vortex shedding flowmeter in a pipeline, wherein said vortex shedding flowmeter further comprises a transducer producing electronic signals proportional to the shedding vortices and pipeline noise;

bbb) means for inputting said transducer signals into an electronic processor;
ccc) means for converting said transducer signals from the time domain to the frequency domain.
ddd) means for removing all but the shedding vortex signals in (ccc) by means of cross-correlations;
eee) means for computing the flow from the remaining shedding vortex signal of (ddd).

19. An apparatus for producing an electronic signal proportional to flow, comprising:
i) means for mounting a vortex shedding flowmeter in a pipeline, wherein said vortex shedding flowmeter further comprises a transducer producing electronic signals proportional to the shedding vortices and noise;
ii) means for converting said transducer signals from the time domain to the frequency domain; and iii) means for removing the noise signals from the vortex signals by means of signal signature comparisons.
iv) means for averaging and digitizing the frequency domain signals of step (ii);
v) means for comparing the repetitive low Q signal signature of the vortex signal relative to the random and/or high Q signals of noise as the means of signal signature comparisons of step (iii).
vi) means for using Fourier analysis in step (ii) to convert said transducer signals from the time domain to the frequency domain.
vii) means for using cross-correlational computations in step (iii) for removing the noise signals from the vortex signals.

20. In a method of measuring flowrate of a fluid in a pipeline having extraneous noise, wherein a transducer is used to generate a time domain flow signal having a frequency indicative of the flowrate of the fluid, but where the noise results in noise signals that are superimposed over the flow signal in frequency ranges in the frequency range of the flow signal, the improvement comprising:
transforming the time domain signals into frequency domain signals;
identifying a stable signature characteristic of the frequency domain flow signal that is dependably different than the frequency domain noise signals;
utilizing said signature characteristic of the flow signal to separate the noise signals from the flow signal;
identifying the frequency of the flow signal; and determining the flowrate of the fluid as a function of the frequency of the flow signal.

21. The improvement of claim 20, including the step of identifying the Q
value of the flow signal and comparing it to the Q value of noise signals in frequency and amplitude ranges comparable to the frequency and amplitude range of the flowsignal, and formulating said signature characteristic in terms of a distinctive Q value.

22. The improvement of claim 21, including the step of removing signal components that have Q values substantially different than the Q value of said flow signal.

23. The improvement of claim 22, including the steps of creating a noise template with a Q value similar to the noise signal components, superimposing said noise template on the signal, and using cross correlation techniques to identify and remove the noise components of the signal.

24. The improvement of claim 23, including the steps of creating a flow template with a Q value similar to the flow signal, and using cross correlation techniques to identify the frequency of the flow signal.

25. Flow meter apparatus for measuring the flow rate of a fluid flowing in a pipe, comprising a monolithic stem body adapted for being positioned in the flowing fluid and having a blunt, bluff body surface portion facing the flowing fluid, anarrowed plate portion trailing said bluff body surface portion, a thin wing portion extending at both its upper and lower ends from upper and lower portions of saidstem body, but both the leading and trailing edges of said wing portion being separated from said narrowed plate portion, said upper portion of said stem portion being shaped in a smooth, rounded contour at its edge that faces the fluid flow and contoured smoothly to a converging, slim trailing edge;
a cylindrical shroud mounted on and encircling the bluff body, narrow plate, and vane portions of said stem body;
an elongated probe extending from the upper portion of said stem body and adapted for insertion through the wall of said pipe; and transducer means associated with said wing portion for sensing flexures of said wing portion and creating electrical signals indicative of said flextures.

26. The flow meter apparatus of claim 25, including signal processing means for determining flow rate of the fluid flowing in the pipe from frequency of said wing flexures.

27. The flow meter apparatus of claim 26, wherein said signal processing means removes noise components from the signal by identifying signature characteristics of the flow signal that are different than the noise signals and using said signature characteristic to remove the noise signals.