JPS6274105A - Function generation method for servo control system - Google Patents

Abstract

PURPOSE:To shorten the arithmetic time, and to offset a terminal condition error by making moving distances at the time of acceleration and at the time of deceleration equal to each other, generating a speed at the time of deceleration by the same calculating equation as that at the time of acceleration, and adding a stepwise step to a deceleration curve. CONSTITUTION:A real time processor is provided with a microprocessor (MPU)20a, a program memory 20b and a RAM20c, and executes an operation processing for generating a function, etc., by executing a program. In this regard, the RAM20c is provided with a parameter area PMA and a work area WA. In this case, a moving distance at the time of acceleration and a moving distance at the time of deceleration are equalized so that an acceleration curve and a deceleration are equalized so that an acceleration curve and a deceleration curve become similar shapes, and a speed at the time of deceleration is generated by the same calculating equation as that at the time of acceleration. Also, a stepwise step is added to the deceleration curve, and a terminal condition error is offset. In this way, a correct terminal condition can be satisfied, and a function can be generated at a high speed.

Description

[Detailed description of the invention] 〔table of contents〕 overview Industrial applications Conventional technology The problem that the invention aims to solve Means for solving problems (Figure 1) Actions Example (α) Description of one embodiment (FIGS. 2, 3, and 4).

(Figs. 5 and 6) (6) Description of other embodiments (Fig. 7) Effects of the invention [Summary] In a function generation method for a servo control system that generates a positioning trajectory of a servo-controlled object as a target position, The time when the difference between the final target position and the current position becomes equal to the moving distance during acceleration is the starting point for deceleration, and by generating a deceleration curve similar to that during acceleration and adjusting the deceleration curve in a stepwise manner, the final target This is to satisfy the termination condition at the position.

[Industrial application field]

The present invention relates to a function generation method for a servo control system that generates a positioning trajectory in real time for performing servo control, and more particularly to a function generation method for a servo control system that improves generation of a positioning trajectory during deceleration.

In recent years, the development of servo control technology for servo-controlling actuators such as motors has been remarkable, and especially in robots and the like, higher speed and higher precision are required.

In addition, digital servo control, particularly software servo using a processor, is being actively developed to replace conventional analog servo. Digital servo control uses a processor as hardware and performs servo control using software, so in addition to being able to easily respond to changes in the characteristics of the servo-controlled object, it also enables easy adjustment and accurate control, and has become popular in recent years. .

[Conventional technology]

Generally, in a digital servo control system, as shown in FIG. 8, a real-time processor 2 receives servo on/off commands and a final target value (final target position) rf from a host processor 1 that controls the entire servo control system. Operation t2L(A) for control is calculated by calculation l, and D/A
It is converted into an analog quantity by a (digital/analog) converter 3 and applied to the servo controlled object 4. The servo controlled object 4 is a power amplifier 4α motor 4
h(!: Consists of encoder 4C, power amplifier 4
The motor 4h is current-driven by the analog operation t u (k) given by a, and the rotation of the motor 4h is detected by the encoder 4C. The output of the encoder 4C is counted by the counter 5, and the observation t y indicating the current position is
(k) is input to the real-time processor 2.

Such a real-time processor 2 includes a function generating unit 20 that sequentially generates a target position γ (A) as a positioning trajectory from a final target position "f," and an error between the target position r (k) and the observed quantity y (A). A certain error a (A) = r (
A) From the error calculation unit 21 that calculates -y (A) and the error e (k), servo calculation is performed using the optimum regulator theory and an observer, and the manipulated variable u (A) by the servo calculation is sent to the D/A converter 3. It has a function of a servo calculation section 22 that outputs.

In the function generating section 20, the target speed curve shown in FIG. 8 (6) is obtained, and the target position r (k> shown in FIG. 8 0) is calculated from the target speed and output as a trajectory.

For such a target speed curve, when accelerating, it is enough to increase the speed to the maximum speed according to the maximum acceleration, but when decelerating, it is necessary to satisfy the boundary condition (terminal condition) that the speed becomes zero at the same time when the final target value is reached. It is necessary to do so. Conventionally, this deceleration curve has been determined by solving an equation expressing the trajectory under the aforementioned boundary conditions, and by performing a factorial calculation where the exponent is a fraction. For example, the trapezoidal velocity curve shown in FIG. 9(g) was determined by square root calculation.

[Problem that the invention seeks to solve]

However, it requires complex calculations such as factorial calculations, and especially when approximate calculations are required, even more complex calculations are required.
There was a problem that high-speed calculation was difficult.

Furthermore, due to errors caused by approximation, there has been a problem in that fine adjustments are required to satisfy the boundary conditions.

SUMMARY OF THE INVENTION An object of the present invention is to provide a function generation method for a servo control system that allows a positioning curve during deceleration to be generated by simple and high-speed calculations in the same way as during acceleration, and which accurately satisfies the termination conditions. .

[Means for solving problems]

FIG. 1 is a diagram explaining the principle of the present invention.

Generally, in order to generate a function using a computer, a function (velocity 2 position) is generated discretely for each sampling.

In Figure 1 (g), the circles are sampling points,
In the present invention, basically the acceleration curve and the deceleration curve have similar shapes. In other words, the moving distance rα during acceleration and the moving distance during deceleration? ``Create the speed during deceleration using the same calculation equation as when accelerating, making h equal to .

Therefore, the deceleration starting point is the point dp' when the movement distance rα during acceleration matches the difference between the final target position "f" and the current position.
Ideally, this would satisfy the termination condition.

However, as described above, because of the discrete function generation, it is rare for the coincidence time point dpt to coincide with a sampling time point of a fixed time interval. That is, it is conceivable that the speed at the acceleration end point deviates slightly from the commanded speed, or that the deceleration start point dp/ cannot be found at the sampling time as shown in FIG. 1(c).

Therefore, the deceleration start point is set as a sampling point dp close to the actual coincidence time dpt, that is, a sampling point where the difference between the final target position "f" and the current position and the moving distance r(z) can be considered to be equal.

For this reason, a termination condition error gr corresponding to the area of the shaded part in FIG.
By adding a step-like step to the deceleration curve as shown in B), the termination condition error δr is canceled out, and the termination condition is satisfied.

[For production]

In the present invention, since a deceleration curve similar to an acceleration curve can be used, the calculation equations for speed and position during deceleration can be the same as during acceleration, making it possible to generate simple and high-speed functions. Since the terminal condition error caused by discrete function generation is offset by providing a step-like step in the deceleration curve, it is possible to accurately satisfy the terminal condition.0 [Example] (α) One implementation DESCRIPTION OF EXAMPLE FIG. 2 is a block diagram of a real-time processor used in the present invention.

In the figure, 20α is a microprocessor (hereinafter referred to as MPU), which performs arithmetic processing for function generation etc. by executing programs, and 20b is a program memory, which stores control programs for arithmetic processing etc. of MPU 20α. 20C is a random access memory (hereinafter referred to as R, AM) that stores parameters, data, etc. necessary for processing of the MPU 20α, and a parameter area PMA that stores parameters; Work area WA for storing data, 20d
is a bus, which includes an MPU 20α and a program memory 20
3 and 4 are flowcharts of function generation processing according to an embodiment of the present invention, and FIG. 4 is a correction processing flow diagram in FIG. 4. FIG.

■ When the MPU 20α receives a startup instruction and a final target position "f" etc. from the host processor 1, the MPU 20α sets the sampling time k to [σ] and the initial position in the parameter area PMA of the RAM 20c. r(0) is "rα," final position %rf, initial velocity v(o) is "0", maximum velocity is ν□.

Set the maximum acceleration to αm1 and the acceleration α to αm.

■ The MPU 20α next determines the target position r at time (A+1).
(k-1-1) is calculated.

First, the speed Zl(A+1) at time h+x is expressed as v(k+
1) = v (&) 1'r -a =-==・(
1) However, T is calculated based on the sampling interval.

That is, the MPU 20α reads out υ(k) and acceleration α of the work area WA of R and AM 20C, performs equation (1), and stores them in the work area WA as v (k-1-1).

Then, the target position y(k-)-1) at time (A+1) is similarly accessed from RAM20 (?), and the following formula 7 %(2
) Calculated by

(2) The MPU 20α looks at the status flag 8F of the AM 20 (?) and determines whether it is in acceleration 9, constant speed, or deceleration.

Incidentally, the state flag SF is set to the acceleration state when setting the initial value.

■ When the MPU20α determines that it is in the acceleration state, the RA
M20? The acceleration travel distance rα of the work area WA is τ
α-r(4).In other words, the target position r(&-1-1)ζ which occurs at time (k+1) versus the target position r(A) which occurs at time Adopt as a position.

(2) Next, the MPU 20α determines whether the current position r(&) is the starting point of the constant velocity state.

That is, it is determined whether the following inequality holds between the speed v (&) at time k and the speed υ(k+i) at time (A+1).

υ(k)50m〈υ(A+1) ・・・・・・・・・
(3) If this inequality holds true, the maximum speed will be exceeded at time (k+1), and it will be assumed that a constant velocity state has started. If this inequality does not hold, it will be determined that acceleration continues.

■ When the MPU 20a determines that equation (3) holds true, it starts the constant velocity state and sets the state flag SF in the RAM 20C.
Set to a constant velocity state, and set the parameter area P' M
Set the acceleration α of A to “0...”, return to step ①, and calculate the velocity v(A+1) and target position r(k+l) at time (&+1) using equations (1) and (2) again. Redo the calculation.

■ In step ■, if the accelerated state where equation (3) does not hold is continued, or if it is determined in step ■ that the state is a constant velocity state based on the state flag SF, the MPU 20σ is set at the target position r (A:
) is the start point of the deceleration state.

That is, the MPU 20α determines whether the following inequality holds.

r f −r (k)≧ra > r f−r(k+
1) ---(4) This means that MPU20α is RAM2
0 C parameter area PMA final target position rf,
Movement distance ra1 target position r during acceleration of work area WA
(k).

r(A-1-1) is read and executed.

If this inequality holds, (final target position - current position)
is equal to the moving distance rα, and it is determined that the deceleration starting point has been reached at the time.

On the other hand, if the inequality does not hold, it is determined that the deceleration starting point has not been reached, and the constant velocity state or acceleration state is continued.

That is, the MPU 20α updates the sample time k and returns to step (2).

(2) When the deceleration start point is reached in step (2) and it is determined that the deceleration state has started, the MPU 20α sets the sign of the acceleration α in the parameter area PMA of the RAM 20C to a negative value, that is, =αm, and sets the status flag SF to the deceleration state.

■ Next, the MPU 20a determines the error δ from the ideal target position.
Calculate r using the following formula.

δr; rf-ra-r(k) (5)
That is, the MPU 20α accesses the RAM 20C, reads the final target position rf, the current position r(k), and the moving distance fa during acceleration, executes equation (5), and stores δr in the RAM 20C.
C, and then return to step (2) and calculate the velocity v(k-)1) and target position at time (A+1) using equations (1) and (2) again. ``Redo the calculation of (A+1).

[Stock] If the deceleration state is determined from the state flag SF in step (2), the process moves to the flow shown in FIG. 4.

First, the MPU 20α determines whether or not to correct the target position.

That is, M P TJ 20 a is R, AM20 /17
) Ware f rear WA v<k”), v (A+t),
δr is read and it is determined whether integers n and n' that satisfy the following equation exist.

v(k), n, T<δr<U(k+1)−
n, 'T ------- (6) The existence of an integer n, , B' that satisfies this equation (6) means that it does not satisfy the terminal condition and that, based on the current speed v (#) This is a good thing as deceleration causes an error δr at zero speed.

On the other hand, the fact that n, n,' that satisfy Equation (6) does not exist means that an error δr does not occur even if the speed is decelerated from the current speed v (k).

■n2 that satisfies equation (6) in step ■. n,'
If it is determined that the target position exists, the target position at time (k+1) is corrected, and the velocity ? Perform uniform motion at J (k).

That is, the MPU 20a first detects the deviation from the ideal target position (
Error) δr is corrected using the following formula.

δr=δr−T, v(k) ・・・・・・・・・・・・
-(7) Also, update the integer n to (tt-1).

Furthermore, MPU20α is RAM201? Set the acceleration α of the parameter area PMA to [3J to make it a constant velocity state, set the state flag SF to a constant velocity state, and return to step ■.
k+1) time velocity v (h11) and target position r
Redo the calculation of (A+1).

■ On the other hand, in step O, satisfy Equation (6).

If it is determined that M does not exist, there is no need for correction, and M
The PU 20α determines whether the final target position "f" has been reached.

That is, MPU2Qcz is tJ (k+1)=O or r(
Check whether A+1)≧”f.

If it is determined that the sample time k has not been reached, the sample time k is updated and the process returns to step (3) to continue deceleration.

(2) If it is determined in step (2) that the final target position has been reached, the MPU 20α corrects the final target position as shown in FIG.

That is, the MPU 201Z has v (A+1)=o and r(k
+i) Check to see if <rf (indicated speed is zero and has not reached the final target position).

If ν(A+1)=0 and r(#+1)<rf, correction settings are made, and function generation is ended.

On the other hand, if u (A+1)=0 and r(k+1)<rf, then the MPU 20α has υ(k+1)>0 and τ(k+i
)≧rf (the commanded speed is positive and the target position r(A-1-1) exceeds the final target position "f").

If ν(i+x)>Q and r(A knee 1-1)≧”f, a correction is set, and the function generation is ended.

Also, when υ(A+1)>0 and τ(#+1)≧rf, Naina et al., 14 (A+1) =0, f (k+1)=ff
Therefore, function generation ends without correction.

To summarize the above operation, at time k, (+)
During acceleration, equation (1) is set as acceleration α-am in step ■.

Calculate v (k + 1) and r (k + 1) using equation (2), update the travel distance τα during acceleration in step ■, determine whether it is the start of constant velocity in step ■, and if it is the start of constant velocity, step ■
Then, v (k+x)−v(A) is adopted as the constant speed.

(11) Also, when the speed is constant, in step
Calculate t' (A+t) and τ(A+1) using the formula, and determine whether it is the time to start deceleration in step ■. If deceleration has started, k
The time is used as the deceleration start point.

01D Furthermore, when decelerating, calculate the error δr in step ■, set the acceleration α = −(L 771 in step ■, and calculate v(k+1
).

r (A+1) is calculated, and further, through step 0 and step (2), the speed u Ck+1) and interval of the step-like steps for canceling the error δr are calculated. That is, when formula (6) is satisfied, the velocity of v(h) is maintained in steps.

翰 To such (1) (++ Even though the calculation was made according to the procedure of
(k+1)=O, if r(k+1)<rf, then (8th)
If u (k+1)≧0 and r (A+1)>”f, then the equation (9) is corrected.

With such simple calculations, it is possible to generate a value on a curve that includes a constant motion approximately equal to the commanded speed, and to reach the target position at zero speed.

Note that the speed v0) determined in this way and the target position r (
A) (where k=o, 1, 2.) is MPU 20 a or another MP at each sampling time.
Servo calculation etc. are performed by U, and the manipulated variable u, (/c
) is applied to the D/A converter 3 (FIG. 8).

Figure 6 shows a speed curve generated by the processing flow, where the speed starts at time (step) no°, deceleration starts at time (step) t, constant speed is corrected at time (t+i), and deceleration occurs again. Then, final error correction is performed.

In this way, the equations during acceleration, equations (1) and (2)
Using the formula, constant velocity, velocity during deceleration, and target position can be obtained.

(h) Description of Other Embodiments FIG. 7 is an explanatory diagram of another embodiment of the present invention, showing an example in which an arcuate curve is used instead of a trapezoidal curve.

In such a case, it can be realized in the same way by simply changing the equation.

Although the present invention has been described above using examples, the present invention can be modified in various ways according to the gist of the present invention, and these are not excluded from the present invention.

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to generate a function value during deceleration using a simple procedure similar to that during acceleration without calculating a positioning curve during deceleration using complicated calculations. is simple and reduces the time required for calculation.

Further, even if a function value is generated during deceleration using the same procedure as during acceleration, a predetermined termination condition can be satisfied, and highly accurate positioning can be maintained.

Therefore, it greatly contributes to increasing the speed and accuracy of the servo control system.

[Brief explanation of drawings]

Fig. 1 is a diagram explaining the principle of the present invention, Fig. 2 is a configuration diagram of an embodiment used in the present invention, Fig. 3,
4 and 5 are flowcharts of function generation processing according to one embodiment of the present invention. FIG. 6 is an explanatory diagram of the operation of one embodiment of the present invention. FIG. 7 is an explanatory diagram of another embodiment of the present invention. The figure is a block diagram of a digital servo control system, and FIG. 9 is an explanatory diagram of the prior art. In the figure, 2... real-time processor, 4... servo control system, 20... function generator.

Claims (1)

[Scope of Claim] A function generation method for a servo control system in which a target speed is determined for each sampling from a given final target position as a positioning trajectory of a servo-controlled object, and a target position is generated from the target speed, comprising: Compare the difference between the target position and the current position and the moving distance during acceleration, and set the sampling point at which the difference and the moving distance are equal as the deceleration start point to perform the deceleration based on a deceleration curve similar to that during the acceleration. A function generation method for a servo control system, characterized in that a target position is generated and the target speed is adjusted in a stepwise manner so that a termination condition at the final target position is satisfied.