JPS60160404A - Automatic tracking system of inertial fluctuation in servo system - Google Patents

Abstract

PURPOSE:To attain initial setting of a gain in a short time by adding an adaptive loop comprising a model and an adaptive device to a control system so as to adjust the gain. CONSTITUTION:The 2nd order IP control system as a feedback speed control system is used, and a model 1 representing a desired response, a feedback control system 2 of a control variable and the adaptive device 3 are provided. The speed command omega* is inputted to both said model and a plant 2, and after a difference omega of both outputs is processed by a positive real compensating element, the result is multiplied with a second order differentiating signal of the plant 2. The result is given to an integration element of a memory and a proportional element K2, and the value added with the initial value K(0) of K is fed back to the plant 2 as the gain value of the K so as to form the adaptive loop. Thus, the desired speed response insensible to the inertia change and external disturbance of load is obtained.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a parallel model reference application system (rMR
ASJ=Model Reference Adap
It is called tiveSVS tem. ) is related to an automatic tracking system for inertia fluctuations in a motor speed control system.

[Prior art and its problems]

Currently, as a feedback speed control method, a method is well known in which an operation such as proportional integration is applied to the difference between a command value and an output to reduce the deviation from a target value. Among these control methods, PI control is the most common, and is approximated by the block diaphragm of FIG. 1, and its input-output relationship, expressed by a transfer function, has the following quadratic form.

Note that the symbols in FIG. 1 are as follows.

CKN: PI control gain, TN: integral time constant ■: current command conversion constant, KT: ) torque constant, J: inertia, 1 (71 speed feedback gain, ω car: speed command, I*: current command, I: armature current, τe: generated torque, τL = torque disturbance, ω: speed] Therefore, in this system, when the gain KN and inertia J of the PI control section are variable and the others are constants, these two
Depending on these values, the response may become slow or oscillatory.

Normally, to prevent this, the inertia J is
When the value of is known, a method is used in which the value of is set in advance so that the output does not become oscillatory. However, if the load is a load whose inertia fluctuates over a large range, such as control of a robot's arm, and the gain of the PI control section is fixed, then if the gain is set small, the inertia will decrease until the command speed is reached. It requires a long time, and if it is set close to the vibration limit, on the other hand, the inertia increases and an unfavorable response with overshoot is exhibited. This causes problems such as excessive positioning and decreased accuracy.

Further, when cutting an object with a machine tool, there is a drawback that the inertia before and during cutting is different, and furthermore, the stability of the speed system is disturbed due to load disturbance during cutting.

[Purpose of the invention]

The present invention solves these conventional problems and selects a system that does not have overshoot in the output as a model and responds in a short time in the feedback speed control system of a rotating machine, and eliminates the occurrence of inertia fluctuations and load disturbances. The purpose of this is to adjust the gain in the velocity loop of the controlled object so that the controlled object can obtain a velocity output equivalent to the output of the model even when

[Structure of the invention]

The present invention is characterized in that, in a speed control system for a rotating electric machine to which an adaptive loop is added using a parallel type model reference adaptive system, a gain in the speed control system is adjusted according to the amount of variation related to the load. That is.

[Detailed description of the invention]

As a speed control system, the PI control system shown in Fig. 1 is suitable (
However, an example of the configuration of a gain tracking adaptive control system for inertia fluctuation will be described below in an IP control system with a simpler transfer function form.

When an IP control system is expressed in a block diagram, it is generally shown in FIG. 2, and the transfer function between the speed command and speed output of this system can be expressed in the following quadratic form.

Note that K is a proportional gain, )('rct is a feed bank gain of 1, and KTG2 is a feed bank gain of 2.

In equation (2), when only inertia J and gain are considered as variable elements, and the others are constants, each term except S2 is multiplied by the coefficient of /J, so this can be reduced to 32 Since it is treated as a variation only in the coefficient of (
By dividing the numerator and denominator of equation (2) by /J, equation (2) can be replaced with the following form.

c G"=a s 2+ b s + c'------
-----'-("'Equation Therefore, when the inertia changes, the only coefficient that changes in Equation (3) is a. Here, Equation (3) is the transfer function of the model in parallel MRAS, and the controlled object (hereinafter referred to as the plant) is expressed by the following formula G p (sl = K c /' (fa s 2
+ b s + c ) ----------141 formula However, K, b, and c are the same as formula (3).

Since the purpose of control is to make the output of the plant equal to the output of the model, the final goal is for 5 to match a, that is, to increase or decrease K at the same rate as the inertia J changes. .

Below is an example of the configuration of an adaptive loop based on Bopov's hyperstability theory, which is often used.

First, let the output of the model be ωH1, the output of the plant be ap, and the difference between the respective outputs ωH−ωp be e, then (
The following differential equation is obtained from equations (2) and (3).

afi+be+ce+ (a-a)ap =0-(51
Equation 5 is selected as the sum of a proportional element and an integral element that acts as a memory, and is set in the following form.

・−・・−・−−−−−−−−−−−−1−−−・−−
−−−−−〜−−(Formula 61, however, ψ! (shi, t, τ),
ψ2(v, t) is a certain function, 50TO) is the initial value of 5,
At the time before τbat, ■ is the output of the true and real compensator C, and ■
can be expressed by the following formula.

v = C (31・e −−−−−−−−−−−−−
−−−−−−−−−−−−−−−(71 formula (5)) Linear element aM+be+ce=−all compensation element
v = Cis)・e We can consider the equivalent feedback system shown in Figure 3, which includes the nonlinear element ωs = (a-5)ip ((1
(1) is a linear element, (12) is a compensation element, and (13) is a nonlinear element), ψ that satisfies Povov's superstability condition for the loop! , ψ2 is determined.

From Bobov's stability condition, 7”v (a-5) 6) p dt ψ2 (v, t) + Ao (ol) ] O p d
t≧−γo2・−−−−・−−−−−・・−・・・・
−1−−−・−・−1−−−−−・−m−−−・(7)
where γ0 is a translation constant In order to satisfy formula (7), the following '2 condition must be satisfied:
−・−−−−−(Formula 81! vψ2(v, t) &p
dt ≧-γ2” −−−−−−−−−−− (Equation 91 where γ1.T2 is a positive constant Equation (8) and the following function is selected as ψ1.ψ2 that satisfies Equation (9).

ψ1 (V+L+τ)=klvωp −・−1−−
−・−−−1−−−αψFormula ψ2 (v, t) = k
2 v &p -------------, - (11) where kl <Q, k2 (constant of Q.

Here, according to Bobov's stability condition, the forward linear element, including the compensator, must satisfy the guest condition, so assuming the transfer function of the compensator to be C (31 = C1s + 02), the true real condition is Determine c, , c2 so as to be satisfied.

Regarding compensation for load disturbance, the sum of the torque command and load disturbance can be treated as inertia and transformation, so the above design can be used as is.

The operation of the entire control system will be explained below using the drawings.

Fig. 4 shows an example in which a second-order IP control system is used as a feedback speed control system. The part surrounded by a dashed line in system (3) shows the adaptation mechanism. ill and (3) can be realized using a microprocessor, etc. 0 line is a speed command that is applied to both the model and plant systems. After inputting and processing the difference ε between the respective outputs using the true-real compensation element, the product is multiplied by the signals for the two similar machines in the plant.The result is converted into the proportional element 2.
and the initial value R(
The adaptive loop is completed by feeding back the value added together with 01 to the plant as the gain value of K.

Figures 6 and 7 compare the output and gain of the conventional fixed gain system and this system.
l, (bl, lad in Fig. 7, (bl correspond to the inertia step change in Fig. 5 (al) and the inertia ramp change in (bl), respectively. The command gives a three-stage step signal and a ramp signal. The solid line represents the model output, the dashed line represents the plant output, and the two-dot chain line represents the change in gain.

The above is an example of the configuration of the adaptive control system in the tp control system, but the adaptive system for the pr control system can also be derived using the same procedure, and its structure is the same as that of the adaptive system shown in FIG.

〔Effect of the invention〕

As described above, according to the present invention, a desired speed response that is insensitive to changes in inertia and load disturbances can be achieved by adding an adaptive loop composed of a model and an adaptive mechanism to a conventionally used control system. This has the advantage that it is possible to perform the control that is obtained, and that the initial setting of the gain can be performed in a short time.

[Brief explanation of drawings]

Figure 1 is a speed control block diagram of the PI control system, Figure 2 is the I
A speed control block diagram of the P control system, FIG. 3 is a block diagram explaining an equivalent feedback system including linear elements, compensation elements, and nonlinear elements, FIG. 4 is a block diagram showing an embodiment of the present invention, and FIG. A time chart showing changes in inertia. Figure 6 is a time chart showing the response when there is no compensation element when a 3-stage speed command is given. Figure 7 is a time chart showing the response when a 3-stage speed command is given. It is a time chart showing a response when an element is added. (1): Speed control system model (2): Speed control plant (3): Adaptive loop patent applicant Yaskawa Electric Co., Ltd. Agent Tebori
Masu (1 idiot) Figure 1 Figure 3 Figure 5 Figure 4

Claims (1)

[Claims] 1. A speed control system for a rotating electric machine to which an adaptive loop using a parallel model standard adaptation system is added, characterized in that the gain in the speed control system is adjusted according to the amount of variation related to the load. Automatic tracking method for inertia fluctuations in servo systems.