JPH0791283A - Engine fuel injection control device - Google Patents

Abstract

(57) [Summary] [Purpose] The deviation of the high-frequency wall flow correction amount and the deviation of the low-frequency wall correction amount are obtained separately, and the learning value corresponding to each deviation is rewritten using each deviation. Thus, the accuracy of rewriting the two learning values is improved. [Configuration] During transition, two wall flow correction amounts Kl and Kh retrieved from the four storage devices 32 to 35 and two learning values Wl
And Wh, the calculation means 40 corrects the basic injection amount Tp to calculate the fuel injection amount during the transition. The calculating means 43 and 44 calculate the absolute value ΔAF of the difference between the peak value of the lean side deviation and the peak value of the rich side deviation in the transient section on the basis of the actual air-fuel ratio calculated from the detected value of the oxygen concentration. The absolute value Δt of the difference between the total deviation time on the lean side and the total deviation time on the rich side in the transient section is calculated as the deviation equivalent amount of the correction amount and as the deviation equivalent amount of the low frequency wall flow correction amount. , The learning values Wl and Wh corresponding to the respective amounts corresponding to the respective deviations are rewritten
5 rewrite each.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an engine fuel injection control device.

[0002]

2. Description of the Related Art Looking at changes in the air-fuel ratio during a transition (provided that the air-fuel ratio during a transition is not corrected), the influence of a slowly flowing fuel wall flow while adhering to an intake port or intake valve ( When a part of the injected fuel is accelerated, it is deprived as a fuel wall flow and the amount of fuel flowing into the cylinder decreases.
When decelerating, the amount of evaporation from the fuel wall flow increases and the amount of fuel flowing into the cylinder increases, so the air-fuel ratio shifts to the lean side during acceleration, and shifts to the rich side during deceleration.

The behavior of such a fuel wall flow is not uniform, and it can be divided into a fuel wall flow having a relatively fast time constant and a fuel wall flow having a relatively slow time constant. There is a method in which the wall flow correction amount is divided into a high frequency component having a fast time constant and a low frequency component having a slow time constant to perform air-fuel ratio correction during a transition, and a learning value is introduced in consideration of a change over time (Japanese Patent Laid-Open No. Sho-06-2006). 63-41634).

Explaining this, the fuel injection pulse width Ti for the fuel injection valve provided in the assembly portion of the intake manifold is Ti = (Tp + Kl.Wl + Kh.Wh) .alpha. + Ts (1) , Tp; Basic injection pulse width Kl; Low frequency wall flow correction amount Wl: Learning value of low frequency wall flow correction amount Kh: High frequency wall flow correction amount Wh: Learning value of high frequency wall flow correction amount α: Air-fuel ratio feedback correction coefficient Ts: invalid pulse width

In equation (1), two wall flow correction amounts Kl
And Kh are values predicted from the operating conditions determined by the engine speed Ne, the injection valve air amount (engine load equivalent amount) Q _AINJ , the cooling water temperature Tw, etc., and during acceleration, as shown in FIG. 26. In addition, the high frequency wall flow correction amount Kh is mainly provided in the initial stage of acceleration, and the low frequency wall flow correction amount Kl is mainly provided thereafter.

Both the two learning values Wl and Wh in the equation (1) are stored in a rewritable map using the engine speed Ne and the cylinder air amount (engine load equivalent amount) Q _CYL as parameters. It is read by referring to the two maps of Wl and Wh.

Regarding rewriting of the two learning values Wl and Wh, first, R = (TFBYA.alpha.) / AFBYA (2) where TFBYA; target fuel air ratio AFBYA; actual fuel air ratio The shortage ratio RTR in the transient state is calculated by using the shortage ratio R in (1) in the following equation: RTR = {R (Tp + Kl + Kh) -Tp} / (Kl + Kh) (3).

Two wall flow correction amounts Kl and Kh and a reference value K _T
Comparing with, the wall flow correction amounts Kl and Kh are both reference values K
_{If T} is exceeded (between A and B in FIG. 26), the high-frequency wall flow correction amount has a deviation from the required value (the deviation of each wall flow correction amount from the required value is simply referred to as a deviation). Then, WH _NEW = Wh _OLD · {1+ (RTR-1) · X _KAT } (4) However, WH _NEW : Wh WH _OLD after rewriting; Wh X _KAT before rewriting; When the map of the learning value Wh with respect to the wall flow correction amount Kh is rewritten and Kh ≦ K _T and Kl> K _T (between B and C in FIG. 26), it is determined that the low frequency wall flow correction amount has a deviation. Wl _NEW = Wl _OLD · {1+ (RTR-1) · X _KAT } (5) where Wl _NEW ; Wl Wl _OLD after rewriting; Wl before rewriting, learning for low frequency wall flow correction amount Kl The map of the value Wl is rewritten.

For example, in FIG. 26, AB at the initial stage of acceleration
If the air-fuel ratio becomes lean during this period, RTR> 1 and the learning value Wh is rewritten to a larger value, and B-C in the latter stage of acceleration is increased.
If the air-fuel ratio becomes rich during this period, the learning value Wl is rewritten to a smaller value than RTR <1.

Considering that the fuel control is performed aiming at the target air-fuel ratio, and finally the supplied fuel amount is obtained from the detected value of the air flow rate, (air flow rate) × (fuel air ratio) = (supply The fuel-air ratio is used because the fuel-air ratio is easier to handle than the air-fuel ratio because the relationship of (fuel amount) is established.

[0011]

By the way, in the above device, the deviation from the target air-fuel ratio during the transition is separated into the deviation of the low frequency wall flow correction quantity and the deviation of the high frequency wall flow correction quantity. Since they are not detected, there is a limit in improving the accuracy of the two learning values together. As shown in FIG. 26, although the high-frequency wall flow correction amount Kh is dominant between A and B in the initial stage of acceleration, the low-frequency wall flow correction amount Kl is also given. Even if there is a deviation in the above, the learning value Wl for the low frequency wall flow correction amount will not be rewritten in the above device.

The deviation of the high-frequency wall flow correction amount Kh is
Before the air-fuel ratio correction by the high-frequency wall flow correction amount Kh is completed, it must be reflected in the learning value Wh for the high-frequency wall flow correction amount. However, in the air-fuel ratio sensor installed in the exhaust pipe, it takes time (there is a response delay) until the exhaust gas reaches the sensor unit and the sensor unit detects the oxygen concentration and outputs it as the actual air-fuel ratio. When the deviation of the high-frequency wall flow correction amount is detected due to the response delay of the fuel ratio sensor, only the air-fuel ratio correction by the low-frequency wall flow correction amount Kl is performed after the end of the air-fuel ratio correction by the high-frequency wall flow correction amount Kh. In some cases, the deviation of the high frequency wall flow correction amount is reflected in the learning value Wl for the low frequency wall flow correction amount (that is, erroneous learning).

Further, in the above device, since the learning values Wl and Wh are rewritten at a constant cycle even during the transition,
It is not known whether the air-fuel ratio correction was performed correctly when the whole transient was seen. The two learning values Wl and Wh cannot be rewritten with high precision unless the deviation of each wall flow correction amount is judged by checking how much the air-fuel ratio deviates from the target value in the entire transition.

On the other hand, the peak value of the deviation in the transient section (see FIG. 27) and the difference or ratio between the area of the lean side area and the area of the rich side area (see FIG. 28) in the transient section are targeted. There is also a method of detecting or calculating the deviation from the air-fuel ratio, but these methods do not distinguish between the deviation of the high frequency wall flow correction quantity and the deviation of the low frequency wall flow correction quantity in the first place. .

Therefore, according to the present invention, the deviation of the high-frequency wall flow correction amount and the deviation of the low-frequency wall flow correction amount are separately obtained, and the learning value corresponding to each deviation is rewritten using each deviation. Therefore, it is intended to improve the accuracy of rewriting the two learning values together.

[0016]

As shown in FIG. 1, a first invention is a means 31 for calculating a basic injection amount Tp according to an operating condition signal, and a low frequency wall changing with a relatively slow time constant. Flow correction amount Kl and high-frequency wall flow correction amount Kh that changes with a relatively fast time constant are stored in association with operating condition signals, respectively, and learning values for the wall flow correction amounts Kl and Kh. Wl and Wh are the wall flow correction amounts Kl, Kh
The devices 34 and 35 for storing the same operating condition signals corresponding to the respective parameters, and the respective wall flow correction amounts Kl and Kh corresponding to the operating condition signals from the four storage devices 32, 33, 34 and 35 at the time of transition. Means 36, 37, 38, 39 for retrieving the learning values Wl and Wh, respectively, and the basic injection amount Tp are corrected by the retrieved two wall flow correction amounts Kl and Kh and the two learning values Wl and Wh. Means 40 for calculating the fuel injection amount during transition, a device 41 for supplying the fuel injection amount to the intake pipe, a sensor (air-fuel ratio sensor or O ₂ sensor) 42 for detecting the oxygen concentration in the exhaust gas, The absolute value ΔA of the difference between the lean side peak value ΔAFl and the rich side peak value ΔAFr in the transient section based on the actual air-fuel ratio calculated from the detected oxygen concentration value.
Means 43 for calculating F (= | ΔAF1−ΔAFr |) as the amount corresponding to the deviation of the high frequency correction amount, and the lean side total in the transient section based on the actual air-fuel ratio calculated from the detected value of the oxygen concentration. Absolute value Δt (= | tl _TA −tr _TA |) of the difference between the deviation time tl _TA and the total deviation time tr _TA on the rich side
In the storage unit 34 for the learning value for the low frequency wall flow correction amount based on the shift amount equivalent to the low frequency wall flow correction amount. Means for rewriting the stored learning value Wl and the learning value Wh stored in the storage device 35 for the learning value for the high-frequency wall flow correction amount based on the shift equivalent amount of the high-frequency wall flow correction amount. And 45.

The second aspect of the present invention, as shown in FIG. 29, means 31 for calculating the basic injection amount Tp according to the operating condition signal.
And a low frequency wall flow correction amount Kl that changes with a relatively slow time constant
And devices 51 and 52 for respectively storing the high-frequency wall flow correction amount Kh that changes with a relatively fast time constant corresponding to at least the temperature of the fuel adhering portion in the intake pipe, and the wall flow correction amount Kl.
A device 53 for storing learning values Wl and Wh for Kh and Kh, respectively, corresponding to the same parameters (at least the temperature of the fuel adhering portion in the intake pipe) as the wall flow correction amounts Kl and Kh.
54, a sensor 55 for detecting the temperature (for example, valve temperature) of the fuel adhering portion in the intake pipe, and the four storage devices 51 at the time of transition using the detected value of the fuel adhering portion temperature,
Means 56, 57, 58, 59 for retrieving the wall flow correction amounts Kl and Kh and the learning values Wl and Wh corresponding to the fuel adhering portion temperature from 52, 53 and 54, respectively, and the two wall flows thus retrieved. Means 40 for calculating the fuel injection amount at the transition by correcting the basic injection amount Tp with the correction amounts Kl and Kh and the two learning values Wl and Wh; and a device 41 for supplying the fuel injection amount to the intake pipe. A sensor (air-fuel ratio sensor or O ₂ sensor) 42 for detecting the oxygen concentration in the exhaust gas, and a peak value for the lean side deviation in the transient section based on the actual air-fuel ratio calculated from the detected value of this oxygen concentration. The absolute value of the difference between ΔAFl and the peak value ΔAFr of the rich side deviation ΔAF (=
| ΔAF1−ΔAFr |) as a deviation equivalent amount of the high frequency correction amount, and a total deviation time tl on the lean side in the transient section based on the actual air-fuel ratio calculated from the detected value of the oxygen concentration. Total deviation time tr between _TA and rich side
Means 44 for calculating the absolute value Δt (= | tl _TA −tr _TA |) of the difference from _TA as the amount corresponding to the deviation of the low frequency wall flow correction amount.
And a storage device 53 for the learning value for the low frequency wall flow correction amount based on the shift amount equivalent to the low frequency wall flow correction amount.
And the learning value Wh stored in the storage device 54 for the learning value for the high-frequency wall flow correction amount is rewritten based on the deviation equivalent amount of the high-frequency wall flow correction amount. Means 45 and are provided.

A third aspect of the invention is the same as the second aspect of FIG.
As shown in 0, Wl _S for two fuels with different fuel properties (e.g. standard gasoline and heavy gasoline), the low-frequency wall flow correction amount at each learning value (standard gasoline for each wall flow correction amount , Wh _S for high-frequency wall flow correction amount for standard gasoline, Wl _H for low-frequency wall flow correction amount for heavy gasoline, W for high-frequency wall flow correction amount for heavy gasoline
h _H ) corresponding to the same parameters as the respective wall flow correction amounts, and two storage devices 61 and 62 for respectively storing the detected values of the fuel adhering portion temperature in the intake pipe from the fuel adhering portion temperature sensor 55. A means 65 for judging whether or not the fuel is at a high temperature above a value, and a means 66 for judging whether or not the fuel used has changed based on the deviation equivalent amount of the high-frequency wall flow correction amount when the temperature becomes higher than the judgment result. When the fuel used does not change from the result of this determination, the learning values WI _S and Wh _S for which the learning for the current fuel (for example, standard gasoline) has progressed are performed on the one hand for the learning values WI _S and Wh _S. When the used fuel changes, the learning values Wl _H and Wh _H with advanced learning for the other fuel (heavy gasoline) are retrieved from the storage device 61 of FIG.
Means 67 for retrieving from the other storage device 62 for _H and Wh _H respectively are provided.

In a fourth aspect of the present invention, as shown in FIG. 31, in the second or third aspect of the present invention, at least one of the post-starting increase correction amount and the ignition timing is a controlled amount, and two different fuel properties are used. fuel (for example, standard gasoline and heavy gasoline) control amount adapted to the respective (for standard gasoline M _S, M _H for heavy gasoline) for storing 2
One storage device 71, 72 and a means 65 for judging whether or not the detected value of the temperature of the fuel adhering portion in the intake pipe from the fuel adhering portion temperature sensor 55 is at a high temperature above a predetermined value.
Then, based on this judgment result, when the temperature becomes high, a means 66 for judging whether or not the used fuel has changed based on the amount equivalent to the deviation of the high-frequency wall flow correction amount; A control amount M _S suitable for the used fuel (for example, standard gasoline) is adjusted from the one storage device 71 for this control amount M _S , and to the other fuel (heavy gasoline) when the used fuel changes. Quantity M
And means 74 for searching each _H from the other storage device 72 for the control quantity M _H, provided with means 74 for controlling the engine in this retrieved control variable.

[0020]

In the first aspect of the invention, the learning value is not rewritten in the middle of the transient section, the waiting value is waited until the end of the transient section, and the deviation of the high-frequency wall flow correction amount from the change in the air-fuel ratio over the transient section. And the shift amount of the low frequency wall flow correction amount are separately obtained as the respective corresponding amounts, and the learning values corresponding to the respective shift amount corresponding amounts are rewritten using the respective shift amount corresponding amounts. The rewriting accuracy is improved for both the values Wl and Wh.

Further, even if there is a response delay in the sensor 42 for detecting the oxygen concentration in the exhaust gas, the value corresponding to the deviation of the wall flow correction amount does not change, so the response delay of the sensor 42 The rewriting accuracy of the learning values Wl and Wh does not decrease.

Next, the generation of the fuel wall flow in the intake pipe is strongly influenced by the temperature of the fuel adhering portion. In the second invention, at least the wall flow correction amounts Kl and Kh are adhered to the fuel in the intake pipe. The learning values Wl and Wh are stored in the storage device corresponding to the part temperature, and the wall flow correction amounts Kl, K are also set for the learning values Wl and Wh.
By storing in correspondence with the same parameter as h, in addition to the operation of the first invention, it is possible to perform wall flow correction according to the degree of generation of the fuel wall flow at that time.

In the third aspect of the invention, it becomes possible to specifically give each learning value for each wall flow correction amount for two fuels having different fuel properties (for example, heavy gasoline and standard gasoline), and the fuel used is, for example, When the learning value for heavy gasoline is learned at the timing of switching to heavy gasoline, it can be switched to the search for each learning value, and the fuel adhering part temperature will differ from the moment when it is switched. Therefore, it does not take time for the entire wall flow correction amount to switch to a value suitable for heavy gasoline.

In the fourth invention, for example, when one fuel is used, a dedicated control amount is used for that fuel, and when the fuel is switched to another fuel, the dedicated control amount is used for the fuel after the switching. Therefore, the amount of fuel increase after the start and the advance of the ignition timing can be performed without excess or deficiency when using any fuel.

[0025]

[Embodiment] In FIG. 2, fuel is injected from a single fuel injection valve 4 provided in an intake port regardless of whether the amount is large or small. Done in. The injection amount increases as the injection time increases, and the injection amount increases as the injection time decreases. The richness of the air-fuel mixture, that is, the air-fuel ratio, shifts to the rich side when the fuel injection amount for a fixed amount of intake air increases, and shifts to the lean side when the fuel injection amount decreases.

Therefore, if the basic injection amount of fuel is determined so that the ratio to the intake air amount becomes a constant value, the same air-fuel ratio can be obtained even under different operating conditions. When the fuel injection is performed once per one revolution of the engine, the basic injection pulse width Tp is calculated from the intake air amount at that time and the engine speed with respect to the air amount sucked in one revolution.
Usually, the air-fuel ratio determined by this Tp is near the stoichiometric air-fuel ratio.

The exhaust pipe 5 is provided with a three-way catalyst 6 for treating three harmful components such as CO, HC and NOx discharged from the engine. The three-way catalyst 6 can efficiently process the three components simultaneously only when the air-fuel ratio of the air-fuel mixture supplied to the engine is in a narrow range centered on the theoretical air-fuel ratio. If the air-fuel ratio deviates from this range to the rich side even a little, the conversion efficiency of the catalyst decreases and the CO and HC emissions increase, and conversely, if it deviates to the lean side, a large amount of NOx is emitted.

Therefore, the control unit 21 detects the oxygen concentration in the exhaust gas over a wide range so that the average value of the air-fuel ratio is maintained near the stoichiometric air-fuel ratio at which the capacity of the three-way catalyst 6 can be sufficiently exhibited. The fuel injection amount is feedback-corrected based on the output signal from the fuel ratio sensor 12.

The control unit 21 also divides the wall flow correction amount into a high frequency component having a fast time constant and a low frequency component having a slow time constant to perform air-fuel ratio correction during a transition, and separately for each wall flow correction amount. (Learning value Wl for low frequency wall flow correction amount Kl and learning value Wh for high frequency wall flow correction amount Kh) are introduced.

In this case, if the deviation from the target air-fuel ratio during the transition is not calculated separately for the deviation of the low frequency wall flow correction quantity and the deviation of the high frequency wall flow correction quantity, two learning values are obtained. W
There is a limit to improving the rewriting accuracy of both l and Wh.

As shown in FIG. 3, the target air-fuel ratio (actually, the range between the lean side limit [A / F] l and the rich side limit [A / F] r is set as an allowable range for the target air-fuel ratio. The deviation from the target air-fuel ratio and the deviation time from the target air-fuel ratio can be divided into two categories.

Here, the magnitude ΔAF of the deviation from the target air-fuel ratio is ΔAF = | ΔAF1−ΔAFr | (6) where ΔAFl; the peak value for the lean side deviation ΔAFr; the peak for the rich side deviation The value defined by the value formula and the deviation time Δt from the target air-fuel ratio are Δt = | tl _TA −tr _TA | (7) where tl _TA ; the total deviation time tr _TA on the lean side. ; It is the value defined by the formula of total shift time on the rich side.

In this case, there is a strong correlation (a direct proportional relationship) shown in FIG. 4 between the value of ΔAF and the deviation of the high frequency wall flow correction amount, and the value of Δt and the low frequency wall flow correction. The present inventor has found that there is a strong correlation (which is also a direct proportional relationship) shown in FIG. 5 with the amount of deviation (which is an experimental result). That is, ΔAF corresponds to the deviation of the high-frequency wall flow correction amount, and Δt corresponds to the deviation of the low-frequency wall flow correction amount, respectively, whereby the deviation from the target air-fuel ratio is calculated by the deviation of the low-frequency wall flow correction amount. It was possible to separate the difference between the high frequency wall flow correction amount and the high frequency wall flow correction amount.

Therefore, in the control unit 21,
The renewal values Wl and Wh are not rewritten until the transition section ends, and the magnitude ΔAF of the deviation from the target air-fuel ratio is set as the equivalent amount of the high-frequency wall flow correction amount, and the deviation time Δt from the target air-fuel ratio. Is calculated as the shift amount equivalent to the low frequency wall flow correction amount, and the learning value corresponding to each shift amount is rewritten using each shift amount equivalent amount (using the shift amount equivalent amount of the high frequency wall flow correction amount). The learning value Wh is rewritten with the learning value Wh using the amount of shift of the low frequency wall flow correction amount).

For such control, the signal from the air-fuel ratio sensor 12 is an air flow meter 7 for detecting the intake air amount, the crank angle sensor 10 for outputting a signal for each unit crank angle and the crank angle reference signal, and the water temperature sensor. 11. A signal from a sensor (not shown) for detecting the valve temperature Tb of the intake valve is input to the control unit 21.

FIG. 6 is a flow chart for calculating the fuel injection pulse width, which is executed at regular intervals.

First, in steps 1 and 2, the actual air-fuel ratio [A / F], the engine speed Ne, and the valve temperature Tb from the air-fuel ratio sensor 12 are read.

Here, the valve temperature Tb represents the temperature of the fuel adhering portion in the intake pipe. When it is difficult to directly measure the temperature of the fuel adhering portion, the equilibrium temperature Th of the fuel adhering portion is obtained based on the cooling water temperature Tw and the intake air temperature Ta, as described in JP-A-1-305142. The predicted temperature value Tf of the fuel adhering portion can be obtained as the first-order lag.

As a simple method, it was confirmed that when an initial value at the time of starting is appropriately selected and brought close to the cooling water temperature with a first-order lag, the value shows a change close to the temperature change of the fuel adhering portion. Therefore, the value Tb1 calculated by the following equation can be treated as the predicted temperature value of the fuel adhering portion. In addition,
The calculation of the equation (8) is performed by a routine executed every one second (independent of the routine of FIG. 6). Tb1 = Tb1 _{−1 sec} + (Tw−Tb1 _{−1 sec} ) × Tb1 _h (8) where Tb1; predicted temperature of fuel adhering portion Tb1 ₋₁ sec; Tb1 Tw one second before; cooling water temperature Tb1 h; correction ratio

In step 3, it is checked whether the absolute value of the difference between the air-fuel ratio [A / F] and the theoretical air-fuel ratio [A / F] st (≈14.5) exceeds a predetermined value K. That is, | [A /
The range of F] − [A / F] st | ≦ K is the allowable range.
For example, 0.8 is set to the predetermined value K. At this time, the lean side limit [A / F] l of the allowable range is 15.3 (= 14.
5 + 0.8), allowable rich side limit [A / F] r
Is 13.7 (= 14.5-0.8) (these limit values are used in steps 16 and 23 of FIG. 14 described later).

From this judgment, if the air-fuel ratio [A / F] is within the permissible range, the process proceeds to steps 5, 6, 7, and 8. If it is outside the permissible range, the learning value updating routine is started at step 4. After that (point A in FIG. 3 is the start timing), the process proceeds to steps 5, 6, 7, and 8 and two wall flow correction amounts K
1 and Kh and two learning values Wl and Wh are obtained from the engine speed Ne and the valve temperature Tb by referring to each map.

The two wall flow correction amounts Kl and Kh are both fixed values, and the data matched with the standard gasoline is written in these maps at the time of initial setting. When the map contents of the wall flow correction amounts Kl and Kh are shown in FIGS. 7 and 8, the tendency of the values is the same, the value increases as the valve temperature Tb decreases, and the engine speed Ne increases as the valve temperature Tb increases. The value is getting smaller.

The map of Kl and Kh is a cylinder air equivalent pulse width (engine load equivalent amount, which will be described later) Av.
9 and 10 using tp and valve temperature Tb as parameters.
It can also be assigned as shown in (Second embodiment). At this time, as shown in FIGS. 9 and 10, a value that increases as Avtp increases even at the same valve temperature Tb enters the initial setting. Similarly, the intake air amount can be used instead of Avtp as shown in FIGS. 11 and 12 (third embodiment).

The two learning values Wl and Wh are stored in a rewritable map (created on the RAM). However, the map of the learning values Wl and Wh is also divided by the same parameters (engine speed Ne and valve temperature Tb) as the map of Kl and Kh. The default data is 1.0,
At the advanced stage of learning, the deviation from 1.0 will be a little due to the absorption of variations due to changes over time and individual differences (differences in flow rate characteristics of injection valves, differences in flow rate characteristics of air flow meters, etc.). The learned values Wl and Wh are backed up by a battery so that data will not be lost even after the engine is stopped.

In step 9, the fuel injection pulse width Ti is Ti = (Avtp + Kl · Wl + Kh · Wh) · TF · α + Ts (9) where Avtp: Cylinder air equivalent pulse width TF; Various correction α: Air-fuel ratio feedback correction coefficient Ts: invalid pulse width is calculated.

The expression (9) is a general expression including both steady and transient, and the low frequency wall flow correction amount Kl ·
Wl and the high-frequency wall flow correction amount Kh and Wh are A
It is added to vtp.

The transient state is judged based on the changing speed ΔAvtp of the cylinder air equivalent pulse width Avtp (for example, when ΔAvtp> 0.2 ms / 2 ms, it is judged as a transient time).
To do.

Cylinder air equivalent pulse width Av of equation (9)
tp is Avtp = Tp * Fload + Avtpn _-1 * (1-Fload) (10) where Tp: basic injection pulse width Avtpn _-1 ; previous Avtp Fload; weighted average coefficient, and (10) The basic injection pulse width Tp of Tp = (Qs / Ne) × K # × Ktrm (11) where Qs: intake air amount obtained by linearizing the output of the air flow meter Ne: engine speed K #; basic A constant Ktrm that determines the air-fuel ratio; a constant that is determined from the flow rate characteristics of the injector.

The various correction TFs in the equation (9) are calculated by the following equation: TF = 1 + Ktw + Kas (12) where Ktw: water temperature increase correction coefficient Kas: start-up increase correction coefficient.

The increase correction coefficient Kas after the start of the equation (12)
Is a value that is determined according to the cooling water temperature during cranking and gradually decreases with time immediately after the engine is started. The water temperature increase correction coefficient Ktw is a value obtained from the cooling water temperature by referring to a table. , (12) are well known.

13 and 14 show two learning values Wl and Wh.
Is executed independently of the flow of FIG. 3 after starting in step 4 of FIG.

At the start of updating in steps 11, 12, and 13, the air-fuel ratio [A / F], engine speed Ne, and valve temperature Tb are read, and four variables [A / F] max,
[A / F] Substitute initial values of 0 for min, tl _TA , and tr _TA . Here, [A / F] max is the maximum value of the air-fuel ratio,
[A / F] min is a variable for entering the minimum value of the air-fuel ratio, tl _TA is the total deviation time on the lean side, and tr _TA is the total deviation time on the rich side.

In step 14, it is checked whether the air-fuel ratio is within the allowable range. If the air-fuel ratio is outside the allowable range (that is, | [A / F]-[A / F] st |> K), FIG.
Go to 4.

At step 15 in FIG. 14, the air-fuel ratio [A / F]
And stoichiometric air-fuel ratio [A / F] st are compared to determine whether the deviation from the allowable range is on the lean side or the rich side.

If [A / F]> [A / F] st, then it is outside the permissible range to the lean side.
Proceed to 22.

First, in steps 16 to 19, the maximum value of the air-fuel ratio in the transient section is obtained. In steps 16 and 17, [A / F]> [A / F] l and [A / F]> [A / F]
If it is max, the routine proceeds to steps 18 and 19 to put the value of the air-fuel ratio into the variable [A / F] max, read the air-fuel ratio, and then returns to step 16, where the air-fuel ratio is set to the lean side limit [A / F]. ] By repeating until the timing becomes 1 (= 15.3) or less, the maximum value of the air-fuel ratio in the transient section enters the variable [A / F] max.

In step 16, [A / F] ≤ [A / F] l
If so, the routine proceeds to steps 20 and 21, where the lean side shift time tl is calculated, and this is added to the variable tl _TA . Specifically, looking at the waveform in FIG. 3, since the process proceeds to step 20 at two timings, point B and point C, in step 20, two lean side shift times tl ₁ and tl ₂ are calculated in FIG. 3, respectively. , The sum of these goes into the variable tl _TA .

In step 22, from the maximum air-fuel ratio value [A / F] max in the transient section, ΔAF1 = [A / F] lmax− [A / F] st (13) Find the value.

On the other hand, step 15 to steps 23 to 2
The process proceeds to 9 when the air-fuel ratio is outside the allowable range and is on the rich side.

The air-fuel ratio is outside the permissible range and is on the lean side. Similar to the above case, steps 23, 24,
25, 26 minimum air-fuel ratio [A / F] m during transition
in, and in steps 23 and 29, the minimum air-fuel ratio value [A /
From F] min, the peak value for the rich side deviation is calculated by the formula ΔAFr = [A / F] st- [A / F] min (14).

Further, in steps 23, 27 and 28, the total shift time tr _TA on the rich side in the transient section is obtained. In the waveform of FIG. 3, steps 27 and 2 are performed only at the timing of point D.
8, the rich side deviation time tr ₁ is the variable tr.
Enter _TA . Note that, during deceleration, since the upper and lower sides of the air-fuel ratio waveform in FIG. 3 are inverted, it goes without saying that the sum of the two rich side deviation times enters the variable tr _TA .

Returning to FIG. 13, in step 14, the air-fuel ratio is within the allowable range (| [A / F]-[A / F] s.
When t | ≦ K), the transient air-fuel ratio fluctuation has ended (whether the air-fuel ratio is in the middle of transient fluctuations (section from point B to point E or point D to point F in FIG. 3)). 3 to determine whether it is after the point C), steps 30, 31
And the stoichiometric time (time when the air-fuel ratio is within the allowable range) tst and a predetermined value tsa (for example, 1 s).
ec) are compared. When tst ≦ tsa, it is determined that the transient fluctuation is in progress, the air-fuel ratio is read in step 32, and the process returns to step 14. In FIG. 3, the time between B and E and the time between D and F are calculated as the stoichiometric time tst. However, since the value of this tst is shorter than tsa, it is judged to be in the middle of a transition, but after the point C. At stoichiometric time t
When st is calculated, it becomes longer than tsa, and it is determined that the transition has ended.

If it is determined in step 31 that ttrans> tsa, then the transient air-fuel ratio fluctuation has ended, step 3
Proceeding to 3, 33, the two learning values Wl and Wh are rewritten separately.

FIG. 15 shows a subroutine for rewriting the learning value Wl for the low frequency wall flow correction amount.

The number of revolutions N at that time in steps 41 and 42
select an address corresponding to the e and the valve temperature Tb, add data at that address in a variable Wl _OLD.

[0066] At step 43, the deviation amount corresponding amount of data and low-frequency wall flow correction amount of the variable Wl _OLD (tl _TA -t
r _TA ), Wl _NEW = Wl _OLD + β · (tl _TA −tr _TA ) (15) However, β: rewrite rate (for example, 10%) is used to calculate new data, and the new data (variable) is calculated. Wl
The value of _NEW ) is stored in the address.

As shown in FIG. 16, with respect to the learning value Wh for the high-frequency wall flow correction amount, WH _NEW = Wh _OLD using the corresponding shift amount (ΔAF1−ΔAFr) of the high-frequency wall flow correction amount. + Β · (ΔAF1−ΔAFr) (16) where, WH _NEW ; data after rewriting Wh _OLD ; data before rewriting is rewritten by the formula (steps 51, 52, 53, 5)
4).

FIGS. 17, 18 and 19 are shown in FIGS.
This is a subroutine for calculating each of the three times (lean side deviation time tl, rich side deviation time tr, stoichiometric time tst) used in Step 2.

Since tl and tr are calculated in the same manner, they will be described with reference to FIG.

In FIG. 17, in step 61, the variable Cl
After the initial value 0 is entered in, the clock pulse counting is started, and in steps 62 and 63, [A / F]> [A / F]
Continue counting for l. In step 63 [A /
When F] ≦ [A / F] l, the routine proceeds to step 64 to end the counting, and the result of the counting is put into the variable Cl.

At step 65, the lean side deviation time tl
[Sec] is calculated by the formula tl = Cl / 100 (17). Frequency of clock pulse is 100Hz
Therefore, if the count number Cl is divided by the frequency, it is converted into a time unit.

As shown in FIG. 19, stoichiometric time ts
About t, it is almost the same as tl and tr, except that the count is finished at the timing when the air-fuel ratio is out of the allowable range (steps 83 and 84).

The operation of this example will be described below.

In this example, the learning value Wl is set in the middle of the transient section.
And Wh are not rewritten, wait until the end of the transient section, and then, from the change in the air-fuel ratio over the transient section, the deviation equivalent amount of the high-frequency wall flow correction amount (| ΔAF1-ΔAFr
|) And the low frequency wall flow correction amount equivalent to the amount of deviation (| tl _TA −t
r _TA |) is obtained separately, and the learning value corresponding to each deviation amount is rewritten using each deviation equivalent amount, so that the rewriting accuracy is improved for both the two learning values Wl and Wh. . For example, the learning value Wl is not rewritten with the deviation of the high frequency wall flow correction amount (on the contrary, the learning value Wh is not rewritten with the deviation of the low frequency wall flow correction amount).
Further, even though the shift amount of the low frequency wall flow correction amount and the shift amount of the high frequency wall flow correction amount are calculated, the learning value is not rewritten.

However, in the conventional example, as shown in FIG. 26, since the deviation of the low frequency wall flow correction amount between A and B is ignored, the learning value Wl is not rewritten, and C Although the deviation from the target air-fuel ratio after the point should be accompanied by the deviation of the low frequency wall flow correction amount, the deviation from the target air-fuel ratio after the point C is also ignored and the learning value Wl is rewritten. There is no way to be.

On the other hand, if there is a response delay in the air-fuel ratio sensor 12, in this example, the transient waveform shown in FIG. 3 simply moves to the right side in time without changing the shape of the waveform. The peak value ΔAFl for the shift amount, the peak value ΔAFr for the rich side shift, the lean side total shift time tl _TA , and the rich side total shift time tr _TA do not change.
That is, even if the air-fuel ratio sensor 12 has a response delay, the rewriting accuracy of the learning values Wl and Wh does not deteriorate.

On the other hand, in the conventional example, since the learning value is rewritten at a constant cycle even during the transition, when the deviation of the high-frequency wall flow correction amount is detected due to the response delay of the air-fuel ratio sensor 12, There is a case where it shifts between B and C in FIG. 26, and at this time, the learning value Wl irrelevant to this is rewritten using the deviation of the high-frequency wall flow correction amount.

Further, the wall flow correction amounts Kl and Kh, which largely depend on the temperature of the fuel adhering portion, vary greatly between low temperature and high temperature. Therefore, for example, values matched at low temperature were used at high temperature. In this case, the wall flow correction amount becomes excessive and the air-fuel ratio goes out of the allowable range and becomes rich, but in this example, each map of the wall flow correction amounts Kl, Kh (also for the learning values Wl, Wh) is a valve. If the valve temperatures are allocated in correspondence with the temperature Tb, the wall flow correction amounts Kl and Kh can be given without excess or deficiency even if the valve temperatures differ.

Now, when the fuel is changed, the same valve temperature T
Even in b, the ratio Kl / Kh of the two wall flow correction amounts Kl and Kh is smaller than that of standard gasoline in heavy gasoline (less volatile than standard gasoline), so when using heavy gasoline, it matches the standard gasoline. Kl and Kh
When the map value of is used, the deviation from the target air-fuel ratio is predicted to be larger than when standard gasoline is used (however, the learned value is not considered).

However, when the experiment was conducted, the difference between the wall flow correction amounts Kl and Kh with respect to the standard gasoline was not uniform, and as shown in FIGS. 20 and 21, the high frequency wall flow correction was performed. Regarding the amount of deviation, it was found that when the valve temperature Tb was high, the difference from the standard gasoline was particularly large (this point is emphasized in FIG. 21).

From this fact, by paying attention to the deviation of the high-frequency wall flow correction amount at high temperature, it can be seen whether the fuel used has changed.

Hereinafter, two maps of learning values for the wall flow correction amount are provided for each of the low frequency and the high frequency (four in total), and each map of the post-starting increase correction coefficient Kas and the ignition timing is respectively provided. An example (fourth embodiment) provided with two types of maps of standard gasoline and heavy gasoline will be described.

As shown in FIG. 22, [1] The amount of deviation from the target air-fuel ratio (the amount corresponding to the deviation of the high-frequency wall flow correction amount) ΔAF is a predetermined value N (for example, 1.
0) is exceeded (step 94). [2] Valve temperature Tb is a predetermined value Tn (for example, 80 ° C.)
That is all (step 95). [3] Learning is sufficiently performed for each learning value map for the wall flow correction amount currently in use (step 9
6). If both of the three conditions are satisfied, it is determined that the fuel used has changed from standard gasoline to heavy gasoline, or vice versa.

In this way, when the fuel used changes, in steps 97, 98 and 99 each map of the learning value for the wall flow correction amount, the map of the post-starting increase correction coefficient Kas,
Switch the ignition timing map from the currently used map to the other map.

The air-fuel ratio waveforms before and after the fuel is switched from standard gasoline to heavy gasoline are shown in FIG. 23. The peak value ΔAFl of the deviation to the lean side and the deviation to the rich side before the switching are shown. The peak value ΔAFr of ΔAFr almost matched (see the upper stage), but ΔAFl increased in the transition after switching from that before switching, and thus | ΔAF1−ΔAFr |>
It is N (see bottom).

Further, when the map is switched to the learning value map for heavy gasoline and the fuel is switched to standard gasoline after sufficient learning has been performed on this map for heavy gasoline, ΔAFr is significantly larger than ΔAFl. Then, | ΔAF1−ΔAFr |> N is established and the map is switched again, and the map for which the learning for the standard gasoline has been completed is used again.

Regarding the above-mentioned post-starting amount increase correction coefficient, since the amount of increase correction for heavy gasoline must be larger, the map for heavy gasoline as shown in FIG. 24 shows the characteristics for standard gasoline (shown by broken lines). Becomes a characteristic in which it moves to the upper right as a whole (as an example, a 30% increase line is shown by a solid line). Similarly, since the heavy gasoline has to be advanced with respect to the ignition timing, in the map for the heavy gasoline as shown in FIG. 25, the characteristic for the standard gasoline (shown by the broken line) is generally on the lower right side. The characteristics are moved (the 30 ° BTDC line is shown by a solid line as an example).

In the fourth embodiment, since the fuel properties of the fuels used can be detected, each fuel can have an optimum control map, and the best matching can be performed with each fuel.

For example, as for the wall flow correction amounts Kl and Kh, if the dedicated learning values Wl and Wh are provided respectively, the fuel used is heavy when only one type of map matching standard gasoline is available. Even when switching to gasoline, the deviation from the target air-fuel ratio becomes smaller as the learning progresses. However, the learning values Wl and Wh
Since each map is assigned with the rotation speed Ne and the valve temperature Tb as parameters, it takes time for the entire learning value map to switch to a value suitable for heavy gasoline.

On the other hand, in this example, there are two types of learning value maps (learning value Wl _S for the low frequency wall flow correction amount and learning value W for the high frequency wall flow correction amount for standard gasoline).
h _S , learning value Wl _H for low frequency wall flow correction amount and learning value Wh _H for high frequency wall flow correction amount for heavy gasoline)
Since the map is prepared and the map is switched every time the fuel used is changed, the wall flow correction can be performed based on the learning value map in which the learning for the fuel has advanced from the moment of switching.

Regarding the map of the increase correction coefficient after starting,
If you do not know that the fuel used has been switched and you can only have one type of map, put a larger value than standard gasoline on the map in consideration of when heavy gasoline is used However, a large amount of it becomes unnecessary when using standard gasoline, resulting in poor exhaust performance.

On the other hand, in this example, when the standard gasoline is used, the map dedicated to the standard gasoline is used, and when the heavy gasoline is switched to, the map dedicated to the heavy gasoline is used. Even when using, the amount can be increased after starting without excess or deficiency.

Regarding the ignition timing, it is possible to always perform the ignition timing control adapted to the fuel property. If you can only have one type of map, think of when heavy gasoline is used, and put a value on the advance side rather than standard gasoline on the map, but that much When standard gasoline is used, it prevents the exhaust temperature from rising. On the other hand, in this example, when using the standard gasoline exclusive map that is set to the retarded side relative to the heavy gasoline when using the standard gasoline, the exhaust temperature rises promptly and the activity of the catalyst is accelerated. Therefore, the exhaust performance is improved.

[0094]

According to the first aspect of the present invention, the means for calculating the basic injection amount according to the operating condition signal, the low frequency wall flow correction amount changing with a relatively slow time constant, and the high frequency changing with a relatively fast time constant. A device for storing the wall flow correction amount corresponding to the operating condition signal, and a learning value for each wall flow correction amount corresponding to the parameter of the same operating condition signal as the wall flow correction amount. A device, means for retrieving each wall flow correction amount and each learning value corresponding to the operating condition signal from the four storage devices at the time of transition, and the two wall flow correction amounts and two learning values thus retrieved, A means for calculating the fuel injection amount at the time of transition by correcting the basic injection amount, a device for supplying this fuel injection amount to the intake pipe, a sensor for detecting the oxygen concentration in the exhaust gas, and a detection value for this oxygen concentration Actual air-fuel ratio calculated from Based on the means for calculating the absolute value of the difference between the peak value of the lean side deviation and the peak value of the rich side deviation in the transient section as the deviation equivalent amount of the high frequency correction amount, and from the detected value of the oxygen concentration. A means for calculating the absolute value of the difference between the total deviation time on the lean side and the total deviation time on the rich side in the transient section based on the calculated actual air-fuel ratio as the equivalent amount of deviation of the low frequency wall flow correction amount; , The learning value stored in the storage device for the learning value for the low frequency wall flow correction amount based on this shift amount of the low frequency wall flow correction amount, and the shift value of the high frequency wall flow correction amount. The means for rewriting the learning value stored in the storage device for the learning value for this high-frequency wall flow correction amount based on the quantity is provided. The oxygen concentration of Even in response delay sensor, never rewritten accuracy of each learning value falls.

A second aspect of the invention is a means for calculating a basic injection amount according to an operating condition signal, a low frequency wall flow correction amount which changes with a relatively slow time constant, and a high frequency wall flow which changes with a relatively fast time constant. A device that stores the correction amount corresponding to at least the temperature of the fuel adhering portion in the intake pipe, and a device that stores each learning value for each wall flow correction amount corresponding to the same parameter as each wall flow correction amount. And a sensor for detecting the temperature of the fuel adhering portion in the intake pipe, and using the detected value of the fuel adhering portion temperature, learning of each wall flow correction amount corresponding to the fuel adhering portion temperature from the four storage devices at the time of transition Means for retrieving the respective values, means for correcting the basic injection amount by the retrieved two wall flow correction amounts and two learned values, and calculating the fuel injection amount at the transition, and the fuel injection amount With a device for feeding the pipe A sensor that detects the oxygen concentration in the exhaust gas, and a peak value for the lean side deviation and a peak value for the rich side deviation during the transient section based on the actual air-fuel ratio calculated from the detected value of this oxygen concentration. Means for calculating the absolute value of the difference as a shift equivalent amount of the high-frequency correction amount, and the total shift time on the lean side and the rich side on the lean side in the transient section based on the actual air-fuel ratio calculated from the detected value of the oxygen concentration. A means for calculating the absolute value of the difference from the total deviation time as a deviation equivalent amount of the low frequency wall flow correction amount, and a means for calculating the low frequency wall flow correction amount based on the deviation equivalent amount of the low frequency wall flow correction amount. The learning value stored in the learning value storage device 53 and the learning value stored in the learning value storage device for the high-frequency wall flow correction amount based on the deviation equivalent amount of the high-frequency wall flow correction amount. Rewrite each value Since there is provided a stage, in addition to the effects of the first invention, it is possible to perform the wall flow correction according to the generation degree of the fuel wall flow at that time.

A third aspect of the present invention is a storage device for storing two learning values for the respective wall flow correction amounts for two fuels having different fuel properties, corresponding to the same parameters as the wall flow correction amounts. And a means for determining whether or not the detected value of the temperature of the fuel adhering portion in the intake pipe from the fuel adhering portion temperature sensor is at a high temperature above a predetermined value, and the high frequency wall flow correction when the temperature becomes higher than the result of this determination. The means for determining whether the fuel used has changed based on the amount equivalent to the amount of deviation, and if the fuel used does not change from this determination result, the learning values for which the learning for the current fuel has progressed Each learning value retrieved from the one storage device for values and learned from the other fuel when the fuel used changes, is retrieved from the other storage device for each of these learning values. Since the means for searching is provided, in addition to the effect of the second invention, it becomes possible to give the respective learning values dedicated to the two fuels having different fuel properties, from the moment when the fuel used is switched. Each learned value can be set to an optimum value for the fuel after switching under all operating conditions where the temperature of the fuel adhering portion is different.

The fourth aspect of the invention is to provide two storage devices for storing at least one control amount for the post-starting amount increase correction amount or ignition timing, which control amount is adapted to each of two fuels having different fuel properties. A means for determining whether or not the detected value of the temperature of the fuel adhering portion in the intake pipe from the fuel adhering portion temperature sensor is at a predetermined temperature or higher, and a high temperature wall flow correction amount Means for determining whether or not the fuel used has changed based on the amount equivalent to the deviation, and if the fuel used does not change from the result of this determination, the control amount that matches the current fuel used is stored in one of the memory for this control amount. Means for retrieving a controlled variable suitable for the other fuel from the device and also when the fuel used changes from the other storage device for this controlled variable, and with the retrieved controlled variable. Is provided with the means for controlling the emissions,
In addition to the effect of the second invention or the third invention, it is possible to increase or decrease the amount of fuel after starting and to advance the ignition timing just enough when using any fuel.

[Brief description of drawings]

FIG. 1 is a diagram corresponding to a claim of the first invention.

FIG. 2 is a system diagram of the first embodiment.

FIG. 3 Magnitude of deviation from target air-fuel ratio | tl _TA −tr
_TA | and deviation time from target air-fuel ratio | ΔAF1-ΔAFr
FIG. 6 is a waveform diagram of an air-fuel ratio change during acceleration for explaining |.

FIG. 4 Magnitude of deviation from target air-fuel ratio | tl _TA −tr
FIG. 6 is a characteristic diagram showing a relationship between _TA | and a deviation amount of a low frequency wall flow correction amount.

FIG. 5: Deviation time from target air-fuel ratio | ΔAF1-ΔAF
FIG. 9 is a characteristic diagram showing a relationship between r | and a deviation amount of a high-frequency wall flow correction amount.

FIG. 6 is a flowchart for explaining calculation of a fuel injection pulse width Ti.

FIG. 7 is a characteristic diagram showing map contents of a low frequency wall flow correction amount Kl.

FIG. 8 is a characteristic diagram showing map contents of a high-frequency wall flow correction amount Kh.

FIG. 9 is a characteristic diagram showing map contents of a low frequency wall flow correction amount Kl of the second embodiment.

FIG. 10 is a characteristic diagram showing map contents of a high frequency wall flow correction amount Kh of the second embodiment.

FIG. 11 is a characteristic diagram showing map contents of a low frequency wall flow correction amount Kl of the third embodiment.

FIG. 12 is a characteristic diagram showing map contents of a high frequency wall flow correction amount Kh of the third embodiment.

FIG. 13 is a flowchart for explaining updating of two learning values Wl and Wh.

FIG. 14 is a flowchart for explaining updating of two learning values Wl and Wh.

FIG. 15 is a flowchart for explaining rewriting of a map of learning values Wl.

FIG. 16 is a flowchart for explaining rewriting of a map of learning values Wh.

FIG. 17 is a flowchart for explaining calculation of lean side deviation time tl.

FIG. 18 is a flowchart for explaining calculation of the rich side deviation time tr.

FIG. 19 is a flowchart for explaining calculation of stoichiometric time tst.

FIG. 20 is a characteristic diagram of the deviation of the low frequency wall flow correction amount for different fuels in the fourth embodiment.

FIG. 21 is a characteristic diagram of the deviation of the high-frequency wall flow correction amount for different fuels in the fourth embodiment.

FIG. 22 is a flow chart for explaining control map switching of the fourth embodiment.

FIG. 23 is a waveform diagram showing changes in each air-fuel ratio before and after switching to heavy gasoline.

FIG. 24 is a characteristic diagram showing map contents of a post-starting amount increase correction coefficient for different fuels.

FIG. 25 is a characteristic diagram showing map contents of ignition timings for different fuels.

FIG. 26 is a waveform diagram for explaining the operation during acceleration of the conventional example.

FIG. 27 is a waveform diagram for explaining how to obtain the deviation from the target air-fuel ratio in the conventional example.

FIG. 28 is a waveform diagram for explaining how to obtain the deviation from the target air-fuel ratio in the conventional example.

FIG. 29 is a diagram corresponding to the claim of the second invention.

FIG. 30 is a diagram corresponding to the claim of the third invention.

FIG. 31 is a diagram corresponding to the claim of the fourth invention.

[Explanation of symbols]

 4 injector 6 three-way catalyst 7 air flow meter 8 intake throttle valve 10 crank angle sensor 11 water temperature sensor 12 air-fuel ratio sensor (oxygen concentration sensor) 21 control unit 31 basic injection amount calculation means 32 low frequency wall flow correction amount storage device 33 high frequency wall Flow correction amount storage device 34 Learning value storage device 35 Learning value storage device 36 Low frequency wall flow correction amount search means 37 High frequency wall flow correction amount search means 38 Learning value search means 39 Learning value search means 40 Fuel injection amount calculation means 41 Fuel Supply device 42 Oxygen concentration sensor 43 High frequency deviation equivalent amount calculating means 44 Low frequency deviation equivalent amount calculating means 45 Learning value rewriting means 51 Low frequency wall flow correction amount storage device 52 High frequency wall flow correction amount storage device 53 Learning value storage device 54 Learning Value Storage Device 55 Fuel Adhesion Part Temperature Sensor 56 Low Frequency Wall Flow Correction Amount Search Step 57 High-frequency wall flow correction amount retrieval means 58 Learning value retrieval means 59 Learning value retrieval means 61 Learning value storage device 62 Learning value storage device 65 High temperature determination means 66 Fuel change determination means 67 Learning value retrieval means 71 Control amount storage device 72 Control amount storage device 73 Control amount retrieval means

Claims (4)

[Claims] 1. A means for calculating a basic injection amount according to an operating condition signal, a low frequency wall flow correction amount that changes with a relatively slow time constant, and a high frequency wall flow correction amount that changes with a relatively fast time constant. A device that stores each corresponding to the operating condition signal, a device that stores each learned value for each wall flow correction amount corresponding to each parameter of the same operating condition signal as each wall flow correction amount, Means for retrieving each wall flow correction amount and each learning value corresponding to the operating condition signal from the four storage devices, and correcting the basic injection amount with these retrieved two wall flow correction amounts and two learning values. Means for calculating the fuel injection amount during transition, a device for supplying this fuel injection amount to the intake pipe, a sensor for detecting the oxygen concentration in the exhaust gas, and the actual space calculated from the detected value of this oxygen concentration. Excess based on fuel ratio A means for calculating the absolute value of the difference between the peak value of the lean side deviation and the peak value of the rich side deviation in the crossover section as the deviation equivalent amount of the high frequency correction amount, and the calculated value from the detected value of the oxygen concentration. A means for calculating the absolute value of the difference between the total deviation time on the lean side and the total deviation time on the rich side in the transient section based on the actual air-fuel ratio as a deviation equivalent amount of the low frequency wall flow correction amount; The learning value stored in the storage device for the learning value for the low frequency wall flow correction amount based on the shift amount equivalent to the low frequency wall flow correction amount is used as the shift amount equivalent amount of the high frequency wall flow correction amount. A fuel injection control device for an engine, further comprising: means for rewriting a learning value stored in a storage device for a learning value for the high-frequency wall flow correction amount. 2. A means for calculating a basic injection amount according to an operating condition signal, a low frequency wall flow correction amount that changes with a relatively slow time constant, and a high frequency wall flow correction amount that changes with a relatively fast time constant. A device for storing at least the fuel adhering portion temperature in the intake pipe, and a device for storing each learning value for each wall flow correction amount corresponding to the same parameter as the wall flow correction amount; A sensor for detecting the temperature of the fuel adhering portion in the pipe, and by using the detected value of the fuel adhering portion temperature, the wall flow correction amount and the learning value corresponding to the fuel adhering portion temperature are respectively transferred from the four storage devices at the time of transition. Means for searching, means for calculating the fuel injection amount during transition by correcting the basic injection amount with these two wall flow correction amounts and two learned values that have been searched, and supplying this fuel injection amount to the intake pipe Device and the discharge gas The difference between the peak value of the lean side deviation and the peak value of the rich side deviation during the transient period based on the actual air-fuel ratio calculated from the detected oxygen concentration value. Means for calculating the absolute value of the absolute value as the deviation equivalent amount of the high-frequency correction amount, and the total deviation time on the lean side and the total deviation on the rich side in the transient section based on the actual air-fuel ratio calculated from the detected value of the oxygen concentration. Means for calculating the absolute value of the difference from the deviation time as the equivalent amount of the low frequency wall flow correction amount, and learning for the low frequency wall flow correction amount based on the equivalent amount of the low frequency wall flow correction amount The learning value stored in the storage device for the value, and the learning value stored in the storage device for the learning value for the high-frequency wall flow correction amount based on the deviation equivalent amount of the high-frequency wall flow correction amount. Set up a means to rewrite each The fuel injection control apparatus for an engine, characterized in that the. 3. Two storage devices for respectively storing, for two fuels having different fuel properties, respective learning values for the respective wall flow correction amounts corresponding to the same parameters as the respective wall flow correction amounts, and the fuels. Means for determining whether or not the detected value of the temperature of the fuel adhering portion in the intake pipe from the adhering portion temperature sensor is at a high temperature above a predetermined value, and a deviation amount of the high frequency wall flow correction amount when the temperature becomes higher than this judgment result. A means for determining whether the used fuel has changed based on a considerable amount, and when the used fuel does not change from this determination result, the learned values for which learning for the current used fuel has progressed A method of retrieving from one of the storage devices and, when the fuel used changes, retrieves the learned value having advanced learning for the other fuel from the other storage device for each of these learning values. The fuel injection control device for an engine according to claim 2, characterized in that a and. 4. A storage device, which stores at least one control amount for an increase correction amount after start-up or a control amount for ignition timing and which stores a control amount adapted to each of two fuels having different fuel properties, and the fuel adhering portion. A means for determining whether or not the temperature detected by the temperature sensor of the temperature of the fuel adhering portion in the intake pipe is higher than a predetermined value, and a high temperature wall flow correction amount corresponding to the deviation amount when the temperature becomes higher than this determination result. Means for determining whether or not the used fuel has changed, and if the used fuel does not change based on this determination result, the control amount suitable for the current used fuel is stored from the one storage device for this control amount, A means for retrieving a control amount suitable for the other fuel from the other storage device for this control amount when the fuel used changes, and an engine is controlled by this retrieved control amount. The fuel injection control device for an engine according to claim 2 or 3, further comprising: