JPH0477325B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to an inductive load control device for an induction motor or the like, and particularly to an inductive load control device suitable for controlling an inductive load using current.

[Background of the invention]

The rotation speed of the induction motor is controlled by a thyristor converter. The circuit configuration at this time is shown in FIG.

In FIG. 1, a thyristor converter 1 is composed of a plurality of thyristors 1A, 1B, 1C, and 1D, and an AC power source 2 is applied between intermediate connection points. The inductive load 3 is composed of, for example, a field winding or an armature winding of a DC motor. The thyristor converter 1 has a power conversion function, and actually has the role of converting a DC power source 2 into DC.

The current control circuit 4 has a current command value Ic and a current detector 6.
The deviation from the current feedback signal I _f detected by is taken. In order to make this deviation zero, the gate pulse generation circuit 5
controls the firing of the thyristor of the thyristor converter 1.

In the above configuration, the current control circuit 4 performs, in addition to the deviation calculation described above, a calculation to provide a compensation characteristic based on various constants of the entire circuit system, that is, a compensation calculation. Compensation calculations include proportional compensation, proportional-integral compensation, proportional-integral-differential compensation, etc., and which one is carried is determined by the overall circuit system. Here, the entire circuit system refers to the entire circuit shown in FIG. Other circuit elements can be similarly represented by some transfer function. Therefore, the entire circuit system is
It refers to a system that can be expressed by a transfer function. In particular, it can be considered that the type of compensation to be used is determined by the characteristics determined by the thyristor converter 1, the AC power source 2, and the load 3.

By the way, with an inductive load control device as shown in Figure 1,
Each constant of the compensation calculation of the current control circuit 4, for example, a proportional gain and an integral gain in the case of proportional-integral compensation, is determined at the time of design based on the characteristics of the transfer function of the thyristor converter.

When the transfer function of the thyristor converter is viewed as a first-order lag element, the gain and time constant of the first-order lag element are determined by the voltage of the AC power supply and the resistance and inductance of the inductive load 3. Therefore, the gain and time constant are uniquely determined at the time of circuit design.

However, the actual voltage of the AC power supply and the resistance and inductance of the inductive load are not necessarily the same as the designed values, and generally take values different from the designed values. Furthermore, when the ambient temperature of the inductive load 3 changes (for example, heat generation due to resistance loss when the load is the field winding or armature winding of a DC motor), the resistance value changes, and the resistance value changes depending on the magnitude of the load current. As a result, the value of inductance also changes.

Therefore, it is desirable to set each constant of the compensation calculation to an optimal value according to the characteristics of the controlled object. A conventional example based on this point of view is Japanese Patent Application Laid-open No. 155418/1983 entitled "Method for Estimating Parameters of Power Converter Devices."

This conventional method will be described below.

FIG. 2 shows a waveform diagram for the control device shown in FIG.
The gate pulse generation circuit 5 generates firing pulses at firing angles α ₀ , α ₁ , α ₂ , α ₃ every half cycle. At firing angles α ₀ , α ₂ , ..., thyristors 1A and 1D are fired, and at firing angles α ₁ , α ₃ , ..., thyristors 1B and 1 are fired.
ignite C. The voltage applied to load 3 at this time e _L
has a positive polarity after each ignition pulse, and a negative polarity before this ignition pulse. Furthermore, the load current i is changed to i=i ₀ , i ₁ , i ₂ , . . . for each ignition pulse.
becomes. and exhibits continuous load current. Note that e ₀ is the AC power supply voltage.

In this waveform diagram, the time point α ₀ when the ignition pulse occurs
The current value i from 1 to time α ₁ when the next ignition pulse is generated draws a trajectory determined by the resistance R of the load 3, the inductance L, the peak value E of the AC power source 2, and the angular frequency ω. Therefore, the gain and time constant that affect the characteristics of the thyristor conversion device can be estimated from the load current i.

Now, when an ignition pulse is given at a phase angle α, the load current i at that time is given by the following equation.

Ld _i /d _t + Ri=Esin (ω _t + α) ...(1) This equation holds true only until the next ignition pulse is generated. (same as L and R in block 23 in Figure 6).

Here, if the initial value of the load current is i _p and the firing angle α=α ₀ , then the load current i is becomes. However, =tan ⁻¹ T〓, T〓=ωL/R.

The time t at which i=i ₁ becomes t=(α ₁ −α ₀ +π)/ω (3) with i=i ₀ as a reference. Introducing equation (3) into equation (2), we get becomes. however, It is. Here, the current difference Δi at firing angles α ₀ and α ₁ is On the other hand, if we integrate equation (2) over firing angles α ₁ and α ₂ , we get becomes.

Equations (6) and (7) represent the gain K and the time constant T〓
By solving the simultaneous equations,
The normal gain K and time constant T can be found.

Here, both sides of equation (7) are multiplied by ω, and equation (6) becomes T〓
By _{multiplying by} Here, if the time constant T=L/R=T〓/ω, then this equation becomes ω・I− _ωT・Δi=K(cosα ₀ +cosα ₁ ). By the way, since the angular frequency ω of the AC power source hardly changes, it can be considered to be constant. In addition, the integral value I, current difference Δi, and phase angle α are also
Since it can be determined by measurement, it can be considered a constant. Therefore, this formula can be expressed as a(n)=b(n)・T+c(n)・K when the n-th firing pulse occurs, and a when the n-1th firing pulse occurs. It can be expressed as (n-1) = b(n-1)·T+c(n-1)·K. Therefore, from both equations, T=a(n)c(n-1)-a(n-1)c(n)/
b(n)c(n-1)-b(n-1)c(n)...(8) K=a(n-1)b(n)-a(n)b(n-1)/
b(n)c(n-1)-b(n-1)c(n)...(9) Note that a(n), b(n), and c(n) are expressed by the following formula.

a(n)=ω・I b(n)=ω・Δi c(n)=cosα _o-1 +cosα _o ...(10) Here, I is the interval between (n-1) and n. It is an integral value, and Δi is a change value within the interval between (n-1) and n.

The prior art described in Japanese Patent Application Laid-Open No. 56-155418 described above calculates the integral value I of the load current during one cycle period of the current and previous firing pulses applied to the thyristor converter.
, the difference Δi in the load current between the current and previous ignition pulses in one cycle, and the ignition phase angle α between the previous and current ignition pulses.
From this, find the values of a, b, and c based on equation (10),
The time constant T and gain K are calculated using equations (8) and (9).

Although it is possible to estimate the time constant T and gain K by the method described above, when applied to an actual electric motor, there are problems described below.

When operating a motor, there are a transient state in which the current is sensitized and a steady state in which the current is approximately constant. In a transient state, the denominators of equations (8) and (9) are not small, so there is no problem in terms of calculation accuracy. However, in steady state,
The denominators of equations (8) and (9) become extremely small, and the calculation accuracy becomes extremely poor.

Furthermore, when the current continuous state as shown in Figure 3 changes to the current intermittent state P, or vice versa as shown in Figure 4, the denominator of equations (8) and (9) does not equal 0, so the calculation result is a value with a large error.

Furthermore, in the Gretzz connection, a commutation overlap angle occurs due to the influence of the inductance on the AC side, and if this is ignored, the error will increase; on the other hand, if this is taken into account,
It becomes extremely difficult to solve equations corresponding to equations (8) and (9).

In conclusion, the above conventional examples are limited to applications where parameter estimation in a transient state is important and where the inductance on the AC side can be ignored; otherwise, the error is large and it is difficult to apply in practice. be.

[Purpose of the invention]

An object of the present invention is to provide an inductive load control device that has accuracy in non-transient, so-called steady parameter measurements and can be applied to general applications in which parameters hardly change during operation.

[Summary of the invention]

The present invention makes use of the fact that circuit inductance and current ripple are almost inversely proportional under steady state conditions.The circuit impedance is equivalently determined by measuring the current ripple, and then the circuit impedance is determined by measuring the current ripple. A transfer function is determined, and a compensation characteristic is obtained based on the transfer function. Ignition pulse control is performed based on this compensation characteristic.

First, the principle of the present invention will be explained.

FIG. 5 shows a configuration diagram of a motor control device in which gate pulse control is performed by a microcomputer. This device is functionally the same as the circuit configuration shown in Fig. 1, but a current transformer 6 for obtaining a feedback value is provided on the power input side of the thyristor converter, and the current control circuit 4 is based on a microcomputer. They differ in two points: they are structured as follows. Further, a motor 3 is assumed as a load and has a field winding 3A. This electric motor 3 becomes an inductive load.

The current control circuit 4 includes a rectifier 11 and an AD converter 1.
0, a microprocessor 9, a common bus 7, and a memory 8. The rectifier 11 rectifies the detected feedback current from the current transformer 6, and the AD converter 10 performs AD conversion on the rectified output. Furthermore, this AD converter 10 has a function of detecting whether or not the feedback current is within the current continuity limit.

The microprocessor 9 takes out the processing contents stored in the memory 8 through the common bus 7, and performs calculations and input/output processing of other process signals according to the contents. In the case of current control, the current command value Ic given in advance is compared with the current feedback value I _f obtained through the current transformer 6, multiplied by a predetermined control gain, and the signal is sent to the gate pulse generator 5. Output.

This control gain is a gain for compensation, and in proportional compensation, the proportional control gain itself,
In proportional-integral compensation, for example, when a first-order lag element is used, a transfer function forming the first-order lag element is used. The first-order lag element under this proportional-integral compensation becomes a first-order lag transfer function with a gain and a time constant as constants.
In the following, a case where proportional compensation is used as the compensation format will be explained.

The gate pulse generator 5 generates a gate pulse with a firing angle corresponding to the given signal to the thyristor converter 1.
Output to. As a result, a current flows through the power transformer 2A (which can be considered equivalent to the power source 2), the converter 6, the thyristor converter 1, and the motor 3, and this current is also input through the AD converter 10 to continue control. Ultimately, this current is controlled to be equal to the current command. In the above control, the current movement, that is, the response, is determined by the control gain stored in the memory 8, the gains of the gate pulse generator 5 and the thyristor converter 1, and the resistance and inductance of the circuit. Here, the control gain is
Usually, the resistance and inductance of the circuit are set to satisfy a predetermined response.

FIG. 6 is a control block diagram when viewed from the equivalent circuit of FIG. 5, and is a diagram for explaining the setting process of the control gain Gc(s) mentioned above.

In FIG. 6a, 21 is the control gain Gc (s), 22
are the gain Kp of the gate pulse generator 5 and the thyristor converter 1, 23 is the resistance R and inductance L of the circuit, 24 is the gain Fc for converting the current into current feedback, 25 is the torque coefficient ζφ and the moment of inertia J of the motor 3, 26 is the torque coefficient ζφ of the motor, ω in the figure is the motor speed, E _M is the reverse starting power of the motor, and I _a is the motor current.

FIG. 6b shows the block diagram of FIG. 6a.
When the field current of the motor is zero, blockage ζφ
This is the equivalent block of FIG. 6d in a short time state in which ω = 0 or I _a no energization time is short and J is extremely large so that ω≈0.

FIG. 6c shows the block diagram of FIG. 6b.
This is an equivalent block in a state that can be considered as |R|<<|LS|. Next, we will discuss whether this condition is satisfied or not.

Generally, the electrical time constant L/R of the circuit is L/R=0.02 to 0.04 (sec)...(11), and the cutoff angular frequency ω _c that determines the response of the current control system is generally ω _c ≒ 500 (rad/s)...(12) It is set as large as about. Also, |R|:|LS|=1:|L/R・S|=1:|L/R・jω| ...(13) Therefore, by substituting the above values into these, |R|:| LS|=1:|(0.02~0.04)×j×500|=1:(10~20)...(14) It can be seen that the above-mentioned condition |R|≪|LS| is satisfied.

Here, according to c in FIG. 6, it can be seen that the controlled object is determined by the circuit inductance L, like the block 26.

The control gain of a conventional control device is determined by using the design value of the circuit inductance to determine the control gain Gc.
(s), the circuit inductance often has an error of ±30 to ±50% with respect to the designed value, and this error becomes an error in the target response time of the actual current control system. It is a well-known fact that this error becomes a response error of the speed control system (not described above). Therefore, in order to eliminate this response error in the speed control system, the control gain was adjusted using the actual machine in order to set the response of the current control system to the target response time.
There is a method of determining Gc(s), but it has the disadvantage that it takes a lot of time for adjustment.

In the present invention, the equivalent circuit inductance is determined from the current ripple to obtain the control gain.

[Embodiments of the invention]

The relationship between circuit inductance and current ripple will be explained.

Figure 7 is a circuit diagram for determining the waveform of current i, where e _U to e _W are the phase voltages of the power supply, L _A is the inductance on the AC side, R _A is the resistance on the AC side, and L _D is the DC side. Inductance, R _D is the resistance on the DC side, and E _M is the back electromotive force of the motor. In addition, the above-mentioned L and R
are L=2L _A +L _D and R=2R _A +R _D. Incidentally, these L and R are the same as those of the conventional example (1).

Figure 8 shows the waveforms of each part in Figure 7.
a is the theoretical output voltage waveform of the thyristor at the firing angle α, and shows the case where the current i is continuous. This theoretical output voltage waveform cannot be actually observed because the thyristor output voltage does not include the voltage drop of the AC side impedance. b is the waveform of the current i, which shows the state of the current continuity limit. c shows the gate pulse waveform applied to the thyristor. The current i in the circuit of FIG. 7 is expressed by the following equation.

Here, α: Firing angle of thyristor converter (rad/s) φ=tan ^-1 (ωL/R) (rad/s) ...(16) ω=2πf ...(17) Other variables are As shown in Figure 7.

Here, if we calculate the ratio of R and ωL in the same way as Equation (14), we get R:ωL=1:ω×L/R=1:2π×50Hz×(0.02~0.04)=1:(6.3 〜12.6) ...(18). From this, φ in equation (16) is found as follows: φ=tan ⁻¹ (ωL/R) = tan ⁻¹ (6.3 to 12.6) = π/2.2 to π/2.1 ≒ π/2 (19). Also becomes. Further, when t≪L/R=0.02 to 0.04 (sec), e ^{- t/L/R} ≒1-t/L/R (21). Here, by substituting the simplified formulas (19), (20), and (21) into formula (15), we get i=√2E _a /ωL{(1-t/L/R) sin(π/2
−α−π/3)+sin(ωt+α+π/3−π/2)}−
E _M /R×t/L/R (22), which is the condition for the limit of continuous current, that is, the boundary condition for a steady state.

i(t=0)=i(t=π/3/ω)=0...(23) In order to satisfy Inserting the condition into equation (22), i=√2E _a /ωL{−sinπ/3+sin(ωt+π/3)
} ...(25) (t=0 to π/3/ω) The time t at which i reaches the maximum in equation (25) is t=π/6/ω. As a result, the current ripple shown in FIG. (Current ripple) = √2E _a k/ωL {−sinπ/3+s
in(π/6+π/3)}=0.134×√2E _a /ωL……(27) From the above, L is L=0.134×√2E _a /ω×1/(current ripple)……(2
8) You can ask for more.

Since the relationship between current ripple and L has become clear from the above explanation, one embodiment of the present invention using this relationship will be described in detail below using the drawings.

FIG. 9 is a processing flowchart of the microprocessor for maintaining the state of the current waveform shown in FIG.

The activation timing of the process shown in FIG. 9 is not particularly defined, but in this embodiment, it is activated periodically at a timing synchronized with the gate pulse. In this process, first, a continuous current state is determined in current continuity determination process 51. If the current is not in a continuous state, that is
If NO, proceed to step 52 and perform processing to slightly increase the input signal to the GPG (gate pulse generator 5: see Figure 5), then reset the current continuity limit flag (step 53). Then, proceed to step 56, which will be described later.

If the result of the judgment process in step 51 is TES, that is, the current is in a continuous state, the process proceeds to step 54, where processing is performed to slightly decrease the input signal to the GPG, and then a continuous current limit flag is set (step 54). 55) and proceed to step 56.

At step 56, the GPG input signal increased or decreased at steps 52 and 54 is output to the GPG 5. Upon receiving this, the GPG 5 outputs a gate pulse corresponding to the input signal to the thyristor converter 1 (step 57).

Increase (decrease) GPG input signal in steps 52 and 54
The reason for this is to make the current flow (not flow) more, but in order to improve the judgment accuracy of the current continuity judgment process in step 51, it is better to keep the increase/decrease as small as possible.

Through the above microprocessor processing, the gate pulse is output from the GPG 5 to the thyristor converter 1, thereby maintaining the state of the current waveform shown in FIG. 8.

FIG. 10 is a flowchart for calculating the inductance L from the current flowing in the circuit in the process of FIG. 9, and the activation timing is synchronized with the gate pulse. The content of the flowchart is the first
As shown in Figure 0, the operation is to judge the current continuous limit flag set in the process of Figure 9, and if it is OFF, the current flow is insufficient and the process is terminated, but if it is ON, the 62 So I ₁ in Figure 8 (normally I ₁ ≒
0) and 63 is the electrical angle π/6 (rad/s)
(This is the process corresponding to t in equation (26))
, and work on I ₂ in Figure 8 at 64. At 65, calculate (current ripple)=|I ₂ −I ₃ |, and at 66, find the inductance L from equation (28). Note that equation (28) has
The GPG input signal, that is, the element of firing angle α is not expressed when the motor is stopped (E _M = 0).
This is because α≒π/2 in (Equation (24)
).

FIG. 11 is a flowchart for determining the control gain from the value of L determined in FIG. 10 and the set response time of the control system, and setting the gain. The activation timing is after the calculation of L. Details of the control gain settings will be described later.

Further, the flowchart of FIG. 12 shows another embodiment of the present invention, which is a method of determining the inductance L from the average value of the current at the current continuity limit. This is determined by calculating the current average value from equation (25) (calculating it by integrating the current value between ignition pulses and averaging it by the time between pulses), I _AVE = 1/π/3ω∫〓 ^{/ 3} 〓 ₀ √2E _a /ωL{
−sinπ/3+sin(ωt+π/3)}d _t =3ω/π×√2E _a /ωL{[−sinπ
/3×t〓 ^/3 〓 ₀ + [−cos(ωt+π/3)/ω]〓 ^/
3 〓 ₀ } =3ω/π×√2E _a /ωL{−√3/
2×π/3ω−−1/2/ω+1/2/ω} =√2E _a /ωL{3/π−√3/2}
=0.089×√2E _a /ωL……(29) Therefore, L is as follows: L=0.089×√2E _a /ω×1/I _AVE ……(30)

As a result, L can be determined using the flowchart shown in FIG. The operation in FIG. 12 is to determine the current continuous limit flag set in the process in FIG. 9 at 81, and if it is OFF, the process is terminated, but if it is ON, I _AVE in FIG. 8 is taken in at 82, In step 83, equation (30) is obtained and L is calculated. The control gain is determined from the calculated L using the flowchart shown in FIG.

Since L obtained at 66 in FIG. 10 may vary slightly depending on the timing, it may be better to add processing to obtain the average several times. 83 in Figure 12
The same is true.

The calculation for obtaining the control gain Gc from the inductance L is as follows.

The response due to the transfer characteristics of the current control system is shown in Figure 6c.
When expressed using variable names, that is, expressed using a transfer function, (response) = 1/Gc x 1/K _P x L x 1/Fc (31). From this formula, when Gc is calculated backwards, it becomes Gc=1/(request/response)×1/K _P ×L×1/Fc (32). Here, K _P is the gain (fixed value) of the thyristor converter 1 determined by the power supply voltage, L is the value obtained by equation (28) or (30), and Fc is the current feedback gain. Note that equations (31) and (32) are for the case where the current control compensation method is proportional, but it can also be applied to other compensation methods (eg, integral compensation, proportional-integral compensation, etc.).

FIG. 13 is a flowchart for controlling the generation of gate pulses based on the control gain Gc. This process is repeated at regular intervals. First of all,
100 takes in the current feedback value I _f , and 101 takes in the current command value Ic. Next, find the deviation (Ic−I _f ),
This deviation (Ic-I _f ) is multiplied by the already determined gain Gc (102). The gate pulse generator 5 is controlled to generate a pulse with a firing angle determined from this multiplication result. The gate pulse generator 5 generates gate pulses according to the firing angle according to this control.

The explanation is based on the current continuous limit, firing angle α≒π/2
(rad/s), but the inductance L can also be determined under other conditions. However, the calculation formula may become complicated.

As described above, according to this embodiment, the control gain of the current control system can be automatically set using the value of the circuit inductance calculated from the current ripple, and the time required for adjustment can be significantly shortened. . Here, the time of adjustment is typically the time of trial run adjustment, but it also includes cases where adjustment is performed periodically or aperiodically during an actual driving section.

〔Effect of the invention〕

According to the present invention, the actual inductance can be determined from the causal relationship between the ripple and the inductance, the compensation content can be set based on this inductance, and the ignition pulse control can be performed from this compensation content. Furthermore, because the calculation content is simple, the time required for compensation can be greatly shortened.

[Brief explanation of the drawing]

Figure 1 is a configuration diagram of a conventional inductive load control device.
Figure 2 is a waveform diagram of each part, Figures 3 and 4 are explanatory diagrams of the drawbacks of the conventional example, Figure 5 is an overall configuration diagram of the inductive load control device of the present invention realized by a microcomputer, and Figure 6 is Fig. 5 is a block diagram showing the device using a transfer function, Fig. 7 is a circuit diagram for determining the current waveform, Fig. 8 is a waveform diagram of each part, and Fig. 9 shows the current shown in Fig. 8. FIG. 10 is a processing flowchart for processing one embodiment of the present invention, FIG. 11 is a processing flowchart for determining the control gain, and FIG. 12 is a processing flowchart for processing another embodiment of the present invention. Processing flowchart, 13th
The figure is a diagram showing a processing flowchart for obtaining a firing angle pulse. 2A...Transformer, 1...Thyristor converter,
3...Motor, 4...Current control circuit, 5...Gate pulse generator, 6...Current detector, 7...Shared bus, 8...Memory, 9...Microprocessor, 10...AD conversion 11... Rectifier.

Claims (1)

[Claims] 1. A thyristor converter that is controlled by a firing pulse and supplies current to an inductive load; A current control means that takes in a feedback current value from the input or output of a converter, determines a deviation between the command current value and the feedback current value, and provides a predetermined compensation characteristic for the deviation, and a compensation characteristic of the current control means. ignition pulse generating means that takes in a deviation value corresponding value, controls generation of an ignition pulse to the thyristor converter according to the corresponding value, and generates a predetermined ignition pulse to the thyristor converter; In the inductive load control device, when the impedance 2πfL of the circuit inductance L and the circuit resistance R have a relationship of 2πfL》R when the inductive load control device is viewed from an equivalent circuit, the output current value of the thyristor converter means for calculating the circuit inductance L from the firing angle of the thyristor converter, the power supply voltage to the thyristor converter, and the frequency f of the power supply voltage;
An inductive load control device comprising: means for setting compensation characteristics in the current control means based on transfer characteristics including the circuit inductance when viewed from the equivalent circuit. 2. The inductive load control device according to claim 1, wherein the output current value, which is an element for calculating the circuit inductance, is a ripple value of the output current of the thyristor converter. 3. The induction according to claim 2, wherein the ripple value is a value calculated from the deviation between the output current value at the time of the ignition pulse and the output current at a time approximately 1/2 cycle after the ignition pulse. Sex load control device. 4. The inductive load control device according to claim 1, wherein the output current value, which is a factor for calculating the circuit inductance, is an average value of the output current at or near a limit value between intermittent and continuous output current.