JPH07129247A - Controller for electric hydraulic servo valve - Google Patents

Abstract

PURPOSE:To reduce adverse influence on a servo system in the case of fault occurrence and to obtain high fault detecting capability. CONSTITUTION:A coil current value I and a current reference value Ir are inputted to a subtracter 21 and their current difference DELTAI is found and inputted to a gain correction coefficient calculating circuit 22. This circuit 22 calculates a gain correction coefficient K from an input command C and the current difference DELTAI according to 'Kk=Kk-1+(DELTAI/.C)'. The value DELTAI/G' obtained by dividing the current difference 61 by the stationary gain G of a servo circuit is a command error equivalent value generating Al and the value 'DELTAI/G.C' obtained by further dividing it by the command C is a gain error equivalent value of the command generating the DELTAI. A last coefficient Kk-1 is added to calculate a coefficient Kk which is required at the point of time. This coefficient Kk is inputted to a gain correcting circuit 23 through a K limiter 24 to multiply the command C by the coefficient Km, and the result is outputted to the servo circuit.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an electrohydraulic servo valve used for an actuator of an aircraft.

[0002]

2. Description of the Related Art In electro-hydraulic servo valves, 1
In a multiple valve (hereinafter referred to as a valve) in which one spool is driven by a plurality of coils, the valve is driven by the circuit shown in FIG. In the figure, reference numeral 1 denotes a multiple-type servo valve, which includes a spool 2 and a plurality of, for example, two force motor coils 3a for driving the spool 2.
3b and position sensors 4a, 4b, and coils 3a,
Servo circuits 5a and 5b are connected to 3b. The servo circuit 5a receives the input command Ca of the A channel (Ach) and the position feedback signal FBa from the position sensor 4a.
The coil 3a is driven based on the
The coil 3b is driven based on the input command Cb of the channel (Bch) and the position feedback signal FBb from the position sensor 4b.

However, in the control device configured as described above, the signal system of the A channel (command Ca, servo gain Ga, position feedback signal FBa) and the B channel are caused by the manufacturing error of the device controlling each coil current. It is difficult to make the signal systems (command Cb, servo gain Gb, position feedback signal FBb) equal to each other, and the coil currents Ia and Ib become non-uniform. When the coil currents Ia and Ib are non-uniform, (1) waste of electric power (2) unnecessary heat generation due to force fight (direction of current between coils, that is, direction of generated force is opposite) (3) Deterioration of controllability is said to occur.

In order to suppress the nonuniformity of the coil currents Ia and Ib, a current equalization processing circuit 6 is provided on the input side of the servo circuit as shown in FIG. 5 so as to equalize the coil currents Ia and Ib. There is. The current equalization processing circuit 6 measures the coil current by the coil current measuring device 7 and performs a process of bringing the value I thereof closer to the current reference value Ir from the reference current source 8. The reference current source 8 has an intermediate value (for example, 1
In the case of A, 0.5A and 0A, 0.5A), the average value of all channels, etc. are used. FIG. 5 shows an A channel control system (control system having a servo circuit 5a) for the coil 3a of the servo valve 1. Similar control systems are provided for the other coils 3b, ....

Conventionally, the current equalization processing circuit 6 has the configuration shown in FIG. In this conventional configuration,
The difference between the coil current value I to be homogenized and the current reference value Ir is obtained by the subtraction circuit 11, and a product obtained by multiplying this difference by a certain gain H (s) 12 is used as a bias 14 and an adder 13 Input to the servo circuit and add it to the input command C to the servo circuit.
I try to get closer to r. The value of the gain H (s) 12 is determined by the operating characteristics of the valve, the required degree of homogenization, and the like.

[0006]

The cause of the non-uniformity of the coil current I is the command C, the position feedback signal FB,
This is largely due to the manufacturing error of each channel such as the servo circuit 5. The expression of these errors is generally expressed by Δ = a (gain error)% ± b (null point error) (1). For example, if the command C is voltage information (full scale = ± 10 V) and the error is ΔC = 5% ± 0.05 V (2), the command C between the A channel and the B channel is
The maximum difference is | ΔCa-b | = 10% + 0.1V (3) The coil current I differs by | ΔCa-b | by a difference Δ
I occurs and ΔI = Ga (C × 10% + 0.1V) (A) (4) Here, Ga = Gb, C is the nominal value of the command.

The current difference ΔI is (A) ΔI≈Ga × 0.1 (A) (5) in a region where the command C is small (C≈0), but (A) a region where the command C is large When (C≈ ± 10), ΔI≈Ga (10 × 10% + 0.1) ≈Ga × 1 (A) (6), and considering the entire area of command C, the gain error of command C with respect to ΔI A large impact.

On the other hand, in the conventional current equalization processing circuit 6, the current reference value Ir in FIG. 6 is considered as the coil current value Ib for the B channel, and the equalization processing for the A channel is assumed. At this time, the bias 14 becomes “H (s) ΔI”, and A
It is added to the command Ca in order to bring the coil current value Ia of the channel close to the current reference value Ir.

If the current reference value Ir becomes an abnormal value here (the cause is a failure of the B channel servo circuit 5b, etc.), the A channel coil current value Ia is normal in the A channel servo circuit 5a. However, due to the effect of equalizing the current, the current value I of the failed B channel is
The current is close to b (current reference value Ir). This adversely affects the servo valve 1 to be controlled and the operation of the main cylinder whose hydraulic pressure is controlled by the servo valve 1 (however, the position feedback of the valve, the main cylinder, etc. If it is a servo system to perform,
This effect is temporary. ). In addition, this causes delay in failure detection or failure in an apparatus that attempts to detect a failure in a servo circuit or the like by monitoring a current value.

As described above, in the conventional control device, when the current equalizing process is performed, the bias 14 for suppressing ΔI is added to the command regardless of the size of the command C, so that the failure occurs. However, there is a problem in that the operation of the controlled object deteriorates and the failure detection capability decreases.

The present invention has been made in view of the above circumstances, and paying attention to the fact that the main cause of the current nonuniformity is the gain error among the manufacturing errors, and the command difference equivalent to the current nonuniformity is generated. To provide a controller for an electro-hydraulic servo valve that has a high failure detection capability and has a small adverse effect on the servo system when a failure occurs, by suppressing the command gain correction to perform current equalization processing. With the goal.

[0012]

An electrohydraulic servo valve control device according to the present invention has a structure including a servo circuit which operates in response to an input command, and a target coil current value and a reference current value. A current equalization processing means for converting a difference from the value into a gain error of the input command to correct the input command to bring the target coil current value closer to the reference current value. And

[0013]

In the electrohydraulic servo valve control device according to the present invention, the difference ΔI between the target coil current value and the reference current value is obtained and input to the gain correction coefficient calculation circuit. This gain correction coefficient calculation circuit uses the command C and the current difference ΔI to calculate “K _k = K _k−1 + (ΔI / G ·
The expression C) "to calculate the gain correction coefficient K _K. In this case, G is the steady gain of the servo circuit, K _k-1 is the previously obtained gain correction coefficient, and K _k is the current gain correction coefficient. In the above equation, the value “ΔI / G” obtained by dividing the current difference ΔI by the steady gain G of the servo circuit is the command error equivalent value for generating the current difference ΔI.
The value “ΔI / G · C” divided by is the gain error equivalent value of the command C that generates the current difference ΔI (gain error equivalent value on the command C creation side). This, in addition to the gain correction coefficient K _K-1 at the time of the previous, calculates a gain correction coefficient K _K required at that time. Then, output by the gain correction coefficient K _K to the servo circuit to correct the command C. Coil current equalization processing is performed by the gain correction of the input command C. Note that the gain correction coefficient K _K
Is updated at regular intervals.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a control device for an electrohydraulic servo valve according to the present invention will be described below with reference to the drawings.

(First Embodiment) FIG. 1 is a first embodiment of the present invention and is a block diagram showing a basic configuration of a current equalization processing circuit 20 by correcting a command gain. As shown in the figure, the subtracter 2 calculates the coil current value I and the current reference value Ir.
1 and the current difference ΔI is calculated and input to the gain correction coefficient calculation circuit 22. The command C for the servo circuit is input to the gain correction circuit 23 and the gain correction coefficient calculation circuit 22. This gain correction coefficient calculation circuit 2
2 calculates the gain correction coefficient K _K from the command C and the current difference ΔI by the following equation. K _k = K _k-1 + (ΔI / G · C) where G is the steady gain of the servo circuit, K _k-1 is the previously obtained gain correction coefficient, and K _k is the current gain correction coefficient.

The gain correction coefficient K _K calculated by the gain correction coefficient calculation circuit 22 is input to the gain correction circuit 23 via the K limiter 24, and the command C is used to input the gain correction coefficient K _K.
Is corrected to correct the command C. Then, the corrected command C is sent to the servo circuit in FIG. 4 (or FIG. 5). The K limiter 24 limits the upper limit value and the lower limit value of the gain correction coefficient K _K within an assumed gain error range in manufacturing.

In the above arrangement, the gain correction coefficient K _K
Is 1 in the initial state. A value “ΔI / G” obtained by dividing the current difference ΔI by the steady gain G of the servo circuit 5 is a command error equivalent value that generates the current difference ΔI, and a value “ΔI / G · “C” is a gain error equivalent value of the command C that generates the current difference ΔI. this,
In addition to the gain correction coefficient K _K-1 previously obtained, to calculate the gain correction coefficient K _K required at that time. And
This gain correction coefficient K _K is input to the gain correction circuit 23 via the K limiter 24 corresponding to the assumed manufacturing gain error range, and the input command C receives the gain correction coefficient K _K.
Is output to the servo circuit 5.

By suppressing the command difference corresponding to the current non-uniformity by the gain correction of the input command C as described above, the coil current equalizing process can be surely performed.

(Second Embodiment) Next, a case where the present invention is applied to a triple DDV (Direct Drive Valve) servo circuit shown in FIG. 2 will be described as a second embodiment.

In the second embodiment, the current equalization processing circuit 20 shown in FIG. 1 is realized on a 16-bit microcomputer and the processing cycle is 250 Hz. In the servo valve 1 according to the second embodiment, one spool is driven by triple coils 3a, 3b and 3c. Each coil 3a,
One channel for each of 3b and 3c (Ach for 3a)
However, 3b corresponds to Bch and 3c corresponds to Cch).
Then, only the Ach servo circuit 5a and the microcomputer 30 for the coil 3a are shown, and the feedback loop system for the servo circuit is omitted.

The current Ia supplied from the servo circuit 5a to the coil 3a is taken out through the resistor 31 and the amplifier 32 and scaled at 2.5 V / A (2.5 per 1 A).
After being converted to V), it is converted into digital data by, for example, a 12-bit A / D converter 33, and is taken into the microcomputer 30 of the corresponding channel as software data. The coil current value Iak fetched by the microcomputer 30 is input to the minus terminal of the subtracter 21 and the intermediate value selection circuit 34, and also to the data link unit 35. The data link unit 35 performs data link with the microcomputer 30 of another channel, sends the coil current value Iak of the digital data to the other two channels, and sends the coil current from the other two channels. The values Ibk and Ick are received and input to the intermediate value selection circuit 34.

The intermediate value selection circuit 34 selects an intermediate value among the three-channel coil current values Iak, Ibk, Ick,
The current reference value Irk is input to the + terminal of the subtractor 21.
The subtractor 21 is configured to detect the current reference value Irk and the coil current value Iak.
Gain correction coefficient calculation circuit 22 by obtaining the current difference ΔIk from
To enter. The gain correction coefficient calculation circuit 22, an input command Ck, the gain correction coefficient K _k-1 obtained current difference ΔIk and last determined the current gain correction factor K _k, K
It is output to the gain correction circuit 23 via the limiter 24.
The gain correction circuit 23 corrects by multiplying the gain correction factor K _k for an input command Ck, command C after correction
Is converted into an analog signal by the 12-bit D / A converter 36 and output to the servo circuit 5a.

Next, the operation of the second embodiment will be described. (1) The coil current value Ia of the A channel is 2.5V
After being scaled by A / A, it is converted to digital data by the A / D converter 33, and is converted to A as soft data.
It is taken into the microcomputer 30 of the channel. Similarly, the coil current value I of the other B and C channels
b and Ic are also taken into the corresponding microcomputer 30. (2) The coil current values Iak, Ibk, Ick of each channel fetched by each microcomputer 30 are sent to the other two channels via the data link section 35 and input to the intermediate value selection circuit 34. (3) The intermediate value selection circuit 34 selects an intermediate value among the coil current values Iak, Ibk, Ick of the three channels and sets it as the current reference value Irk. (4) The gain correction coefficient calculation circuit 22 sets the initial value of the gain correction coefficient K _k to “1”, and the following expression K _k = K _k−1 + (ΔIk / Ck · G) where ΔIk = Irk−Iak G = 3.717 (constant value) to obtain the gain correction coefficient K _k . The gain correction coefficient K _k is processed by the K limiter 24 so that 0.9 ≦ K _k ≦ 1.1, and then input to the gain correction circuit 23. The gain correction circuit 23 corrects by multiplying the gain correction factor _{K k} to the input command Ck, through D / A converter 36, and outputs to the servo circuit 5a. Further, the same processing is performed on the other channels, and the gain of the input command Ck is corrected.

The above (1) to (4) are the processing 1 for current equalization.
This is processed at a cycle of 250 Hz, for example. Further, since the gain correction coefficient calculation circuit 22 of (4) includes division by the variable Ck, “| Ck
In the region of | ≦ 0.5 (V) ”, the gain correction coefficient K _k is not calculated and is set to“ K _k−1 ”(the same as the previous correction coefficient calculation value). This is done to prevent overflow, and in the above range of “| Ck | ≦ 0.5 (V)”, the amount of current non-uniformity due to manufacturing error (especially gain error) is within the allowable range. Therefore, this overflow prevention processing becomes possible.

Now, in the second embodiment of FIG. 2, the command Ck = 0 (V), the coil current nominal value (= Ia,
Ib, Ic) = 0 (A), the data link unit 3
It is assumed that 5 fails and Ibk = Ick = 4 (A) [10 (V) after AD conversion].

When the coil currents Ibk and Ick both become 4 (A), the intermediate value selection circuit 34 outputs Irk = (intermediate value of 0, 4, 4) = 4 (A) as the current reference value Irk. . Therefore, the current difference ΔIk output from the subtractor 21 becomes ΔIk = Irk−Iak = 4-0 = 4 (A) and is input to the gain correction coefficient calculation circuit 22.

However, the gain correction coefficient calculation circuit 22
Since the command Ck is 0 (A), " _Kk = _Kk-1 "
Since the gain correction coefficient K _k is calculated as, the Ibk and Ick are “0 → 4 (A)” due to the failure of the data link unit 35.
The effect of being set to does not appear.

In the second embodiment of FIG. 2, the intermediate value selection circuit 34 selects the intermediate value of the coil current values.
The circuit configuration may use an average value instead of the intermediate value.

(Third Embodiment) FIG. 3 shows an example of a third embodiment of the present invention applied to a servo circuit monitoring apparatus. In FIG. 3, the feedback loop system for the servo circuit 5 and the servo circuit model 41 is omitted.

This servo circuit monitoring device has 16 bits
An arithmetic operation of the mathematical model 41 of the actual servo circuit is performed on the microcomputer 30A, and the current value I'k of the model 41 is compared with the actual coil current value Ik to detect the failure of the servo circuit 5. And executes the following processing. (1) The actual coil current value I is scaled at 2.5 V / A, passed through the 12-bit A / D converter 33, and taken into the microcomputer 30A as the coil data value Ik of the soft data. Become. (2) The subtracter 21 calculates the current difference ΔIk between the coil current value I′k of the servo circuit model 41 and the actual coil current value Ik.
Is calculated by ΔIk = Ik−I′k and is input to the servo circuit monitor 42 and the gain correction coefficient calculation circuit 22. (3) the gain correction coefficient calculation circuit 22, | Ck |> When 0.5 _{(V), K k = K} k-1 + (ΔIk / Ck · G) where, G = 3.717 | Ck | ≦ When 0.5 (V), the gain correction coefficient K _k is calculated with K _k = K _k-1 . (4) The gain correction coefficient K _k calculated by the gain correction coefficient calculation circuit 22 is input to the gain correction circuit 23 via the K limiter 24 of 0.93 ≦ K _k ≦ 1.07. The gain correction circuit 23 multiplies the gain correction coefficient K _k to the input command Ck, and inputs the result to the servo circuit model 41 in the next cycle.

As described above, (1) to (4) are equalization processing cycles of the servo circuit model current with respect to the actual coil current of the servo circuit, which are processed at a cycle of 250 Hz.

In this servo circuit monitoring device 40, since the manufacturing error of the actual servo circuit and the error between the theoretical values (= servo circuit model) have an extremely large effect on the coil current values of both, the present invention Current equalization processing is required.

As an application example, K _k = K _k-1 + F (s) × Δ
It is possible to adjust the characteristics by providing a filter having characteristics of Ik / (Ck · G) and F (s).

Now, in the circuit of FIG. 3, assuming that the current is made uniform by using the conventional configuration shown in FIG. 6, H (s) = 7 (V / A) and the bias is ± 0.7 ( V) limit, (0.7 (V) = 10 (V) full scale ×
7%) Command Ck = 0 (V), coil current nominal value = 0
Assuming that the actual servo circuit fails in the state of (V) and the actual coil current I becomes 4 (A), Ik = 4 (A) ΔIk = Ik−I′k = 4-0 = 4 (A) Bias = 7 × 4 = 28 (V) (However, the bias is limited to 0.7 (V) by the limiter), whereby the current I′k output from the servo circuit model 41 is , I′k = (0 + 0.7) × 3.717 = 2.6 (A). Therefore, at this time, the current difference ΔIk input from the subtractor 21 to the servo circuit monitor 42 is ΔIk = 4-2.6 = 1.4 (A).

On the other hand, when current equalization is performed under the same conditions according to the third embodiment of FIG. 3, the gain correction coefficient calculation circuit 22 determines that "K _k = K" because the command Ck is "0".
_{Since the} gain correction coefficient K _k is calculated as “ _k−1 ”, even if K _k−1 = 1.07 (maximum value), the coil current I′k of the servo circuit model 41 is I′k. = Ck × K _k × G = 0 × 1.07 × 3.717 = 0
(Steady value). Therefore, from the subtractor 21 to the servo circuit monitor 42
The current difference ΔIk input to is ΔIk = 4-0 = 4 (A).

Even if a failure occurs under the same conditions as described above, (1) current equalization in the conventional configuration: ΔIk = 1.4 (A) (2) current equalization in the third embodiment of the present invention: ΔIk =
There is a large difference between the input signal ΔIk of the servo circuit monitor 42 and 4.0 (A).

If the filter for the input signal ΔIk of the servo circuit monitor 42 and the current difference level for determining a failure are the same, the current equalization of the present invention can detect the failure in a shorter time. For example, filter 50ms
ec First-order low-pass filter, if the decision level is 1 (A), (1) Conventional configuration 1 (A) = 1.4 (A) × (1-e
^{t / 0.05} ), t = 62.6 msec (2) Third embodiment 1 (A) = 4 (A) × (1-e
^{From t / 0.05} ), t = 14.4 msec, and in the third embodiment of the present invention, it is possible to detect the failure in about 1/4 time as compared with the conventional configuration.

[0038]

As described above, according to the present invention, there is little adverse effect on the servo system when a failure occurs, and high failure detection capability can be maintained. Therefore, control of multiple electrohydraulic servo valves is possible. It is useful for devices and their monitoring devices.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a first embodiment of a control device for an electrohydraulic servo valve according to the present invention.

FIG. 2 is a circuit configuration diagram showing a second embodiment of the present invention.

FIG. 3 is a circuit configuration diagram showing a third embodiment of the present invention.

FIG. 4 is a basic configuration diagram of a control device for an electrohydraulic servo valve.

5 is a configuration diagram showing an outline of a coil current equalizing process for FIG.

6 is a block diagram showing details of a conventional current equalization processing circuit in FIG.

[Explanation of symbols]

 1 Servo Valve 2 Spool 3a, 3b Coil 4a, 4b Position Sensor 5a, 5b Servo Circuit 6,20 Current Leveling Processing Circuit 21 Subtractor 22 Gain Correction Coefficient Calculation Circuit 23 Gain Correction Circuit 24 K Limiter 30, 30A Microcomputer 33 A / D converter 34 Intermediate value selection circuit 35 Data link section 36 D / A converter 41 Servo circuit model 42 Servo circuit monitor

Claims (1)

[Claims] 1. A control device for an electrohydraulic servo valve having a servo circuit that operates according to an input command, wherein a difference between a target coil current value and a reference current value is calculated as a gain error of the input command. The control device for an electrohydraulic servo valve, further comprising: current equalization processing means for converting the input command into a correction value and correcting the input command to bring the target coil current value closer to the reference current value. .