JPH0956162A - PWM controller for converter - Google Patents

Abstract

(57) [Summary] [Construction] The converter 2 is provided with a pattern generator 10 that gives a pulse pattern according to the power supply phase at a constant cycle asynchronously to the power supply, and the point at which the polarity of the power supply voltage is switched is estimated in advance and A means 110 for generating a transition pulse is provided in a cycle period including the polarity switching point so that the pulse pattern is equivalently the same pattern with respect to the power supply voltage in the cycle period in which the points exist and the cycle period subsequent thereto. [Effect] Output pulsation and jump can be suppressed to a minimum even with an inexpensive system in which the smoothing circuit is set small.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to PWM control of a converter, and more particularly to a PWM control device capable of smoothly performing power supply synchronization processing between an AC power supply and a PWM pulse pattern.

[0002]

2. Description of the Related Art Conventionally, synchronization processing between a pulse pattern of a PWM converter and a power supply voltage is disclosed in Japanese Patent Nos. 1877122 and 1878647.
As shown in No. 4, the AC phase value, which is the integrated value of the power supply frequency or its estimated value, is corrected to a new phase value by the elapsed time from the zero cross, and a pulse is generated from this phase value at regular time intervals asynchronously with the power supply. A power supply synchronization method using a non-instantaneous synchronization method with a time delay to calculate the pattern, or a zero-cross as an interrupt is accepted, and a pulse pattern that matches the power supply voltage phase is secured at the minimum pulse width (if the width becomes narrow, There has been proposed an instantaneous synchronization method, etc., which is executed while performing (such as expanding processing).

[0003]

In the first prior art described above, since it is not instantaneous synchronization in essence, when the zero cross input occurs immediately after the scheduling of the latest pulse width, the power supply voltage relationship after the zero cross is met. Since the setting of the pulse pattern is delayed until the next pulse pattern calculation, the PWM pulse pattern that does not match the phase relationship of the power supply voltage after the zero crossing will continue to be output for a short time, and a DC reactor or In an inexpensive system in which a capacitor is not sufficiently inserted, the output of the converter pulsates greatly, and this pulsates through the inverter and flows into the motor as a pulsating current, which causes torque shock and noise.
On the other hand, also in the second conventional technique, when a microcomputer is used for pulse generation, the operation system recognizes the zero-cross interrupt, jumps to the processing program destination that actually generates the corresponding pulse, and completes the processing. Is accompanied by a delay of several tens of microseconds (because of so-called overhead), and there is a problem that the output jumps due to the adverse effect of the processing such as the pulse being extended with the minimum pulse width secured. In such a system, the adverse effect of the discontinuity of the output was a big problem in terms of vibration and noise. Further, these problems are unavoidable in the case of a system using a GTO or a power transistor as a switching element, because the pulse width calculation interval must be lengthened because the switching frequency cannot be set high, and an inappropriate pulse pattern is supplied. The adverse effects become more apparent due to the above-mentioned long period and the like.

It is an object of the present invention to provide a power converter which does not cause pulsation associated with power source synchronization in DC output even in an inexpensive system using a smoothing device having a small capacity.

[0005]

In order to achieve the above object, the present invention provides a converter with a pattern generator which gives a pulse pattern according to the power supply phase at a constant cycle asynchronously with the power supply, and the polarity of the power supply voltage is Estimate the point to switch in advance, and in the cycle period in which the point exists and the next cycle period following it, in the cycle period including the polarity switching point so that the pulse pattern is equivalently the same pattern with respect to the power supply voltage. Means were provided for generating the transition pulse.

[0006]

The transition pulse becomes an ignition signal that minimizes adverse effects on the phase relationship of the power supply voltage, and there is a gap between the pulse pattern and the power supply voltage in the non-instantaneous synchronous pulse system with dead time based on post-processing. There are no problems due to matching, delay in interrupt processing due to the instantaneous synchronization method, or problems due to pulses that are out of the ideal state to secure the minimum pulse width.Therefore, output ripples and jumps appear to be set in a small smoothing circuit. Even an inexpensive system can be kept to a minimum.

[0007]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a system diagram showing an embodiment of the power converter of the present invention. First, the overall configuration will be described. Reference numeral 1 denotes a three-phase AC power supply for supplying electric power to the converter 2, a load 4 is connected to the converter 2 constituted by the switch elements 21 to 26, and the flowing current is detected by the detector 3. The detected current if is fed back as a feedback signal, compared with the direct current command I * input from the current command input terminal 5 at the addition point 6, and the deviation is taken into the pattern generator 10. Also, the total phase creation means 1
In 20, the internally generated power frequency command ω * (50 Hz in the east of Kanto) or the actually detected value is used for a certain time Δt.
The estimated phase PH1 * of the power supply voltage is calculated. The estimated phase PH1 * of the power supply voltage is detected by the voltage detector 7 and taken into the pattern generator 10 via the waveform shaping circuit 8 that generates a pulse at the time of zero crossing that occurs approximately every 60 degrees and the synchronization device 9, The estimated phase PH1 * in the total phase creating means 120 is synchronized. On the other hand, the fetched current deviation Δi causes the function generating means 110 to generate the second phase command PH2 * and the conduction ratio command PU * according to the magnitude thereof. In the total phase generation means 120, the second phase command PH2 * is added to the estimated phase PH1 * subjected to the synchronization process to generate the total phase command θT * which is the basis of pulse pattern generation. The total phase command θT * is input to the area determination means 130, and the total phase command θT * is 0 ° to 60 ° (M1), 60 ° to 120 ° (M2), 120 ° to 180 ° (M3), 180 °. ~ 2
40 ° (M4), 240 ° to 300 ° (M5), 300
The information Mo that indicates which 60 ° section of 360 ° to 360 ° (M6) is included is obtained, and further the total phase command θT
Distributing means 140 for determining what kind of pulse pattern should be given to which element together with *, the duty ratio command PU *
Is input to Then, in the distributing means, when the power source becomes a zero cross, as a pulse pattern, as a dedicated pulse pattern including the zero cross point within the cycle period Δt, a pulse pattern alternating with a pulse pattern immediately after the predetermined cycle period Δt is used. By setting the transition pulse pattern so that the output voltage is equivalently the same with the power supply voltage, it is non-instantaneous with the power supply interrupt, but PWM that prevents output jumps. Control can be realized.

Next, the operation of the distribution means 140 will be specifically described with reference to a flowchart. FIG. 2 shows a basic pulse distribution processing 140. This distribution processing 140 is started every predetermined time Δt, and in processing 14100, the mode information Mo indicating which 60 ° section the total phase command θT * is in, the total phase command θT *, and the flow rate command PU * are read. , Processing 14200, the reference pulses Pd, Pi,
Ps is calculated from the total phase command θT * and the conduction ratio command PU *. Here, Pd is a pulse given to a phase in which the pulse width gradually decreases with increasing phase, Pi is a pulse given to a phase in which the pulse width gradually increases, and Ps is a so-called pulse for a short-circuited phase. Is a reference pulse for. Then, in process 14300, it is determined whether or not it is immediately before the zero crossing, and if Yes, process 144
The transition pulse is set at 00, and if No, process 1
At 4500, it is determined whether or not it is immediately after the zero cross, that is, immediately after the transition pulse is generated.
Set the pulse immediately after transition at 600, and if No,
In processing 14700, the normal pulse is set and the predetermined time Δ
The pulse distribution process for each t is completed. FIG. 3 shows a detailed flowchart of the transition pulse setting process 14400. Process 14
In 401, whether the absolute value of (total phase command θT * -360 °) is smaller than the predetermined value α, that is, the total phase command θT * must be processed to zero cross near 360 ° near 0 °. Determine if it must
If es, the element 2 corresponding to the decrease phase in processing 14402
4 is not ignited, and the pulse width Pd for the element 24 is equally divided into the increasing phase pulse width Pi of the element 25 corresponding to the increasing phase and the short circuit phase pulse width Ps of the element 26 corresponding to the short circuit phase. . The element 23 is an arm on the side that does not commutate (this 6
It is an element that continues to fire for a predetermined time Δt in the 0 ° section). Similarly, the processing 14404 is performed for the transition pulse before 60 °, the processing 14406 is performed for the transition pulse before 120 °, and the processing 14 is performed for the transition pulse before 180 °.
Process 1441 for the transition pulse before 410,240 °
In the transition pulse before 1,300 °, each pulse of the processing 14408 is set as a transition pulse in a predetermined section Δt in which it may exist prior to the occurrence of the zero cross. This transition pulse has a phase relationship between the pulse immediately after transition and the power supply voltage shown in FIG. 4, and forms a pulse pattern that does not cause a jump of the DC output voltage across the zero cross. FIG. 4 shows a flowchart of the pulse setting process 14600 immediately after the shift. In processing 14601, the mode information Mo is M1,
That is, it is checked whether or not the total phase command θT * is within 0 ° to 60 °, and if Yes, the process 14602 is executed.
Then, the pulse generation process immediately after the transition pulse following the transition pulse from M6 to M1 is performed. Specifically, in addition to the original pulse width Pd for the decreasing phase, the element 23 has a pulse width Pi for the increasing phase.
Of the pulse width Ps for the short-circuited phase and the pulse width Pi for the increased phase as well.
The sum of 1/2 is given. Further, the element 25 is always provided with Δt as an ignition for the arm on the non-commutation side. Processing 14402 of FIG. 3 and processing 14602 of FIG.
, The element to which the pulse is applied is different, and the commutation arm is also different (the negative arm is commutated in FIG. 3, whereas the positive arm is commutated in FIG. 4). Since the relationship between and does not collapse even if the zero cross occurs in the middle of the predetermined section Δt, the mode transition is smoothly performed. In addition,
Here, the zero cross from M6 to M1 is described as an example, but the processing 14604, the processing 14606, and the processing 1460 are performed.
8, processing 146010 and processing 14611 also correspond to the zero crosses at every 60 ° and similarly operate as pulse generation immediately after the transition in relation to the transition pulses in FIG. Further, FIG.
In the processing 14401, the method of using the fact that the total phase approaches n * 60 ° is used to estimate the occurrence of the zero cross, but the width of the decreasing phase pulse Pd of the direct reference pulse becomes sufficiently narrow. It is also possible to make a determination by using the fact and the mode information Mo. In addition, in FIG. 3, the example in which the pulse width Pd for the decreasing phase is equally divided into the pulse width Pi for the increasing phase and the pulse width Ps for the short-circuited phase in FIG. 3 is shown in FIG.
i is 2 for the decreasing phase pulse width Pd and the short circuit phase pulse width Ps
Although an example of equal distribution has been shown, the same jump suppressing effect can be obtained even if all are added to the side that is not the short circuit phase pulse width Ps from the viewpoint of power ring. In this case, the division process is reduced, which is advantageous for speeding up the microcomputer. FIG. 5 shows a normal pulse pattern setting process 1 in a section away from the zero cross.
4700 shows a flowchart of 4700. In this case, too, the explanation will be given by taking the example of the situation where the total phase command θT * is in the range of 0 ° to 60 °. In process 14701, it is checked whether or not the total phase command θT * is in the range of 0 ° to 60 °, and if Yes, in process 14702, the normal pulse for M1, that is, the decreasing phase pulse Pd to the element 23, the element 21 A pulse Pi for the increasing phase, a pulse Ps for the short circuit phase on the element 22,
A command is issued to the element 25 of the negative arm that does not perform commutation to continue firing in the predetermined section Δt. In other modes, process 14
By the processing from 704 to 14711, pulse output according to each mode Mo is performed. It should be noted that, here, even if the pulse Pi for the increasing phase is given to the element 21, there is no problem due to the restriction of the minimum pulse width for the following reason. For example, assuming that the switching frequency is 2.7 kHz (Δt = 370 μs), approximately 9 pulses will be included in the 60 ° section, and if the pulse set immediately after transition is set immediately after zero crossing, the pulse immediately after transition will continue. The pulse width Pi of the increasing phase, which is the narrower width of the next normal pulse set generated by about 40 μs (= 370 μs * sin (60 ° / 60 ° /
9)), which is such that the restriction of the switch element can be avoided.

FIG. 6 shows converter components 21 to 26.
An example of a pulse pattern given to is shown. Here, the U-phase zero crosses from negative to positive, that is, the mode information MO is M6.
An example in which a zero cross occurs in the second Δt2 section in the region from M1 to M1 will be described. In the first section of Δt1, the mode information MO is M6, the process 14708 for M6 of the normal pulse setting process 14700 operates, 24, 25 and 26 are commutated, and 23 is always fired.
Here, it is determined that the zero crossing is not close yet, so the normal pulse output is performed, and the calculation of the pulse width and the scheduling of the pulse output are completed before the section of Δt1 is entered. Then, the pulse distribution processing 140 is started again before entering the next Δt section, which is Δt2. Since it is predicted that the zero crossing will come to this section here, the process 14402 of the transition pulse process 14400 is executed instead of the normal pulse pattern for M6, and the following relationship between the pulse immediately after transition and the power supply voltage is shown. As the almost same pulse pattern, 23, 25, 26
The pattern that turns on is output and the output schedule is executed before the zero cross. Here,
Since the pattern in which 25 and 23 are turned on does not cause any problem in relation to the power supply voltage even in the next M1 section, the order of generation is set to follow the short-circuit mode (mode in which 23 and 26 are turned on). This is effective in the relationship of matching between the pattern and the power supply voltage when controlling the expansion / contraction of the length of Δt so that the zero cross is in the latter half of Δt. Next, Δt3
Prior to the section, the pulse distribution processing 140 is activated.
In this case, since the pulse output immediately after the zero-cross has come in this section, 14602 of the process 14600 immediately after the transition is executed, and the process of turning on 22, 23, and 25 is performed.
This pulse pattern is similar to the pattern used for Δt2 in relation to the power supply. Then, when the time further passes and before the Δt4 section is entered, the pulse distribution processing 1
40 is activated. Here, in this section, the pulse output is performed after a while after the zero-cross comes, and therefore the normal processing 1470 is performed.
14702 out of 0 is executed, 21, 22, 23, 2
5 is turned on, and a series of pulse outputs near the zero cross is completed.

FIG. 7 shows an embodiment showing the effect of the present invention. Here, only the pulsating component is taken out and measured in the state where a direct current of 70 A is constantly applied. (A) is a DC current waveform when the conventional control is used, and (b) is a DC output current waveform when the transition pulse and the pulse immediately after transition of the present invention are used. In (a), although the current drop is a little large in the first half where the power supply fluctuation is small, the synchronous processing is relatively stable, but in the latter half, the power supply frequency seems to have fluctuated a little and the optimum power supply voltage is affected. The relationship between the pulse pattern is broken and the current is jumping, whereas in (b), the pulsating current component based on the synchronization processing every 60 ° is well controlled, and almost no jumping motion occurs. I understand. Thus, in the embodiment of the present invention, the pulsation of current can be reduced by 1/3 or more as compared with the conventional method,
Vibration and noise characteristics have been greatly improved. From the opposite viewpoint, this effect shows that the smoothing circuit such as DC reactor can be downsized and the torque pulsation and noise generation can be kept at the conventional level even if the system is made inexpensive, and the effect of cost reduction is shown. It can be demonstrated.
As described above, according to the present embodiment, the power supply synchronization process associated with the zero cross at every 60 ° power supply angle can be smoothly performed, so that there is an effect of reducing the output pulsation.

[0012]

According to the present invention, since the power source synchronizing process associated with the zero cross at every 60 ° electrical angle can be smoothly performed, there is an effect that the output pulsation can be reduced. Moreover, since a smoothing circuit such as a DC reactor can be simplified, cost reduction and device installation area reduction effects can be expected.

[Brief description of drawings]

FIG. 1 is a system diagram of an embodiment of the present invention.

FIG. 2 is a flowchart of an overall control program according to an embodiment of the present invention.

FIG. 3 is a flowchart of an individual control program according to an embodiment of the present invention.

FIG. 4 is a flowchart of an individual control program according to a second embodiment of the present invention.

FIG. 5 is a flowchart of an individual control program according to the third embodiment of the present invention.

FIG. 6 is a timing chart showing an example of pulse output according to the present invention.

FIG. 7 is a characteristic diagram of an experimental result diagram showing the effect of the present invention.

[Explanation of symbols]

1 ... 3-phase AC power supply, 2 ... Converter, 3 ... Current detector,
4 ... Load, 5 ... Current command input terminal, 6 ... Addition point, 7 ... Voltage detector, 8 ... Waveform shaping circuit, 9 ... Synchronization device, 10 ...
Pattern generating device, 21-26 ... Switch element constituting converter, 110 ... Function generating means, 120 ... Total phase creating means, 130 ... Region determining means, 140 ... Distributing means.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Toshio Arita 6-6 Kandanishikicho, Chiyoda-ku, Tokyo Inside Hitachi Building System Service Co., Ltd.

Claims (10)

[Claims] 1. A power converter comprising a plurality of switching elements for exchanging electric power between an AC power source and a direct current, and an AC phase existing at a predetermined cycle which is sufficiently shorter than a half cycle of the AC. In the PWM control device provided with means for determining a pulse pattern from the width of the pulse to be given to the switching element to be generated within the cycle and the switching element to be turned on and off, the PWM control device is
Prior to the arrival of the point at which the polarity of the AC power supply voltage changes in the predetermined cycle, as a pulse pattern for a period including that portion in the cycle, a pulse pattern immediately following the predetermined cycle period and the AC power supply. A PWM control device for a converter, comprising means for setting equivalently almost the same transition pulse pattern in relation to a voltage. 2. The PWM control device for a converter according to claim 1, further comprising means for estimating a polarity change point of the power supply voltage in advance. 3. The polar change point estimating means according to claim 2, wherein the polarity change point is estimated based on information as to whether the AC phase value approaches 60 ° * n (n is an integer). A PWM controller for the converter that estimates it. 4. The PWM control of a converter according to claim 2, wherein the polarity change point estimating means estimates the calculated pulse width before the polarity change based on information as to whether or not the calculated pulse width has become sufficiently narrow. apparatus. 5. The PWM controller for the converter according to claim 1, wherein the power converter is a current type. 6. The transition pulse pattern according to claim 1, wherein:
A PWM control device for a converter configured by a total of two sets of two-phase modulation pulses, one of which corresponds to a short-circuit phase among the three phases, and the other of which corresponds to the wide-side phase of the two-phase power supply phases. 7. The converter according to claim 6, wherein the pulse on the output side of the feeding phase has a width of a predetermined value or more.
WM control device. 8. The PWM control device for a converter according to claim 6, wherein the pulse on the non-output side of the power feeding phase is equally divided into the output power feeding phase and the short-circuited phase. 9. The PWM control device for a converter according to claim 6, wherein all the pulses on the non-output side of the power feeding phase are distributed to the power feeding phase to be output. 10. The PWM control device for a converter according to claim 6, wherein a pulse corresponding to the short-circuit phase of the transition pulse pattern is supplied within a predetermined period prior to a pulse corresponding to another power feeding phase.