JP4220366B2 - Torsional vibration system controller - Google Patents

Description

  The present invention relates to a torsional vibration system control apparatus for controlling a motor of a softly coupled mechanism used in office automation (OA) equipment, factory automation (FA) equipment, and the like.

Conventionally, a torsional vibration system control device that controls a motor of a softly coupled mechanism is known (for example, see Patent Documents 1 and 2).
Patent Document 1 discloses a control device, and Patent Document 2 discloses a torsion system control device based on a state space method.
In addition, in the torsional vibration system control device under development, when the drive side inertial body and the driven side inertial body are controlled objects connected by a torsion shaft, the driven side inertial body to which disturbance is applied is positioned with high accuracy. Although an attempt is made to provide a position control means for controlling, this technique uses the estimated disturbance value as it is for state estimation in real time.
Japanese Patent No. 3266391 Japanese Patent No. 3244184

このような系を古典的な制御手法（ＰＩＤや位相補償等）で制御しようとすると、制御対象の出力１変数のみで制御を行なうため、制御対象の共振周波数において位相遅れの増大が問題となり、これを超えて制御することは困難である。 In a motor drive system, an elastic body element is often included in a torque transmission mechanism, and a resonance system is configured in relation to the moment of inertia of the motor itself or the driven inertial body. At this time, if the rotational accuracy is not so required, the torque transmission mechanism can be regarded as a rigid coupling, so the influence on the control by the resonance system can be ignored.
However, when the rotational accuracy becomes severe, high gain control is required to satisfy the specifications, and the upper limit frequency that must be controlled increases. Such systems are no longer difficult to handle with rigid coupling.
FIG. 5 is a schematic diagram illustrating a physical model of a softly coupled two-inertia system to be controlled. Each symbol in the figure has the following meaning.
· _{J M:} drive-side inertia moment · _{D M:} drive side viscous resistance · _{J L:} driven inertia · _{D L:} driven viscous resistance · _{K C:} torsion shaft spring constants · _{D C:} torsional axis viscous resistance · tau _M : Drive shaft torque (operation amount)
・ Ω _M : Drive shaft angular velocity ・ θ _M : Drive shaft angle (observation output)
・ Τ _L : Driven shaft torque (disturbance)
・ Ω _L : driven shaft angular velocity ・ θ _L : driven shaft angle (observation output)
The resonance frequency ω _{0 at} this time can be expressed by the following equation.
ω ₀ =

When such a system is controlled by a classical control method (PID, phase compensation, etc.), control is performed with only one output variable of the controlled object, so that an increase in phase delay becomes a problem at the resonance frequency of the controlled object. It is difficult to control beyond this.

On the other hand, in the method based on the state feedback control, since all the internal states of the control target are handled, it is possible to control beyond the resonance frequency by performing the state feedback control in a category where the control target can be regarded as a linear system.
In order to realize the state feedback control, it is necessary to know all the state variables. However, there are many cases where the actual control target can observe the state, and the state feedback control cannot be performed as it is.
Therefore, in order to realize the state feedback control, a state estimator for estimating the internal state of the controlled object is prepared. At this time, disturbance torque τ _L is required as an input to the state estimator in addition to manipulated variable τ _M and observation outputs θ _M , θ _L , but in general, measurement of disturbance torque is difficult.
Therefore, for the sake of simplicity, the disturbance torque τ _L is set to zero or a constant value. However, in both cases, a large error occurs in the output of the state estimator due to the difference from the disturbance torque τ _L actually generated. May be destabilized.
FIG. 6 is a schematic diagram showing a configuration in which the twist of the elastic element is generally ignored and treated as a rigid body. As such a disturbance estimator, as shown in FIG. 6, one that generally treats a torsion of an elastic element and treats it as a rigid body is known.
However, in the case of the control target as taken up here, there is a difference in the disturbance torque τ _L (estimated value) estimated from the difference between the elastic body and the rigid body, and there is a risk that the system will become unstable as described above.
Therefore, in view of such a viewpoint, an object of the present invention is torsional vibration provided with a disturbance torque estimator newly considering an elastic body element in order to realize a state estimator with small estimation error used for state feedback control. It is to provide a system control device.

In order to solve the above problems, the invention according to claim 1 is characterized in that a first angle detector is connected to the motor, the motor is connected to the elastic body element, and the second angle detector is connected to the motor. The inertial body to be connected, the motor operation amount value input to the motor, the first angle value detected by the first angle detector, and the second angle detected by the second angle detector A disturbance estimator that calculates a torque value applied to the inertial body using a value, a disturbance memory that stores the torque value calculated by a disturbance estimator, a torque value that is stored in the disturbance memory, and the Using the motor operation amount value, the first angle value, and the second angle value input to the motor, an angle value and an angular velocity value that are internal states of the control target including the motor and the inertial body are obtained. A state estimator to estimate; Based on the deviation between the angle value and angular velocity value, which are the internal state of the controlled object estimated by the state estimator, and the angle value and angular velocity value, which are control target values for the motor and the inertial body, are input to the motor. A torsional vibration system control device comprising a controller for calculating a motor operation amount value .
According to a second aspect of the present invention, in the first aspect of the invention, the state estimator uses a torque value stored in the disturbance memory when the inertial body makes one revolution. It is a state estimator that estimates an angle value and an angular velocity value that are internal states of the controlled object .
According to a third aspect of the present invention, in the first aspect of the present invention, the controller has a resonance frequency at which the control target has an upper limit value of the frequency of the control target including the motor and the inertial body. The controller is controlled to a value higher than the value .

According to the present invention, since the disturbance torque applied to the driven shaft side by the disturbance estimator can be estimated in the softly coupled two-inertia system, the estimation accuracy of the state estimator is improved, and robust and stable feedback control is achieved. It becomes possible.
In addition, the disturbance torque detector and each state detector other than the observation output are not required, so the configuration of the control system can be simplified and the cost can be reduced. Can be stored.
Therefore, in periodic disturbances, disturbances can be removed up to a higher frequency band than when there is no disturbance memory due to a feedback loop. High-precision rotation control can be realized.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of a state feedback control system of a torsional vibration system control apparatus according to the present invention.
In FIG. 1, the object to be controlled is connected to a motor 1 (= drive side inertial body) connected to a drive shaft (not shown) by a torsion shaft (elastic body element) 1a and connected to a driven shaft (not shown). an inertial body 2 (= driven inertial body), the motor 1 and the inertial body 2 each angle detector 3, 4 is connected to the inertial body 2 has a configuration in which the disturbance torque tau _L is applied Yes.
The control unit C includes a disturbance estimator 5, a disturbance memory 6, a state estimator 7, and a state feedback controller 8, and the rotation obtained by the two angle detectors 3 and 4 as an input to the control unit C. The angles θ _M and θ _L and the target state value X _R desired by the designer are input.
A motor drive PWM amplifier 9 is connected to the output τ _M (= operation amount) of the control unit C, and the output of the motor drive PWM amplifier 9 is connected to the motor 1.
FIG. 2 is a detailed view showing the disturbance estimator 5 of FIG. Since the softly coupled two-inertia system is the control target, the disturbance estimator 5 takes into account the elastic body element corresponding to the softly coupled portion, and the disturbance torque τ _L (this τ _L is an estimated value. It is configured to estimate (displayed as (~)) according to the symbols in the drawing. T is a time constant of the low-pass filter.
FIG. 3 is a detailed view showing the state estimator 7 of FIG. Input the manipulated variable, the observed output, and the estimated disturbance torque to estimate all the internal states of the controlled object. The feedback gain E inside the state estimator 7 can be determined by a Kalman filter design method.
However, it is obtained by solving the Riccati equation by giving appropriate noise covariance data. As a design method, since the state estimator has a Kalman filter, the designer can easily design without considering the pole arrangement or the like.
FIG. 4 is a detailed diagram illustrating the state feedback controller of FIG. The manipulated variable τ _M is calculated from the deviation between the estimated internal state X (˜) (ω _M (˜), ω _L (˜), θ _M (˜), θ _L (˜)) and the target state X _R.
The state feedback gain F is determined by an optimal regulator design method, and is obtained by giving an evaluation function appropriately so as to satisfy the control specifications and solving a Riccati equation.
At this time, in order to suppress the resonance of the controlled object, it is necessary to give an evaluation function so that (the upper limit frequency to be controlled)> (the resonant frequency of the controlled object). By using this optimum regulator as a design method, the designer can easily design without considering the pole arrangement or the like.

Next, the operation of FIG. 1 will be described. Respective rotation angle information obtained from the angle detector 3 attached to the shaft of the motor 1 and the angle detector 4 attached to the inertial body 2 is input to the disturbance estimator 5 and the state estimator 7.
Further, the operation amount τ _M output from the state feedback controller 8 is input to the motor drive PWM amplifier 9 and also to the disturbance estimator 5 and the state estimator 7.
The disturbance estimator 5 estimates the disturbance torque τ _L (˜) applied to the driven inertial body 2 from the manipulated variable τ _M and the two rotation angles θ _M , θ _L using the model to be controlled. In this case, the estimation in a manner that takes into account the elasticity K _C and viscosity D _C of the control object inside the flexible linked axial.
The estimated value output from the disturbance estimator 5 is temporarily stored in the disturbance memory 6 without being used in real time. In recording, in the case of the present embodiment, since the driven inertial body 2 is a rotating body, it can be regarded that the applied disturbance torque has periodicity.
Then, after the driven inertial body 2 makes one rotation, that is, when the same rotation angle is reached, the stored disturbance estimated value is used in the state estimator 7 to perform feedforward control. Thereby, about the disturbance which has periodicity, the disturbance suppression effect can be heightened to the high frequency band (about 20-200 Hz) which is a banding band.
The state estimator 7 uses the model of the controlled object from the disturbance torque τ _L (˜), the manipulated variable τ _{M and the} two rotation angles θ _M , θ _L which are estimated in advance, and uses the controlled object model X ( ˜) (ω _M (˜), ω _L (˜), θ _M (˜), θ _L (˜)) are estimated.
At this time, the estimated states θ _M (˜), θ _L (˜) are always compared with the rotation angles θ _M , θ _L. When an error occurs, the estimated values θ _M (˜), θ _L (˜) are adjusted to the rotation angles θ _M , θ _L by the effect of the feedback gain E. To correct the estimated states θ _M (˜) and θ _L (˜).
The state feedback controller 8, and the control target of the internal state X (~), with respect to the deviation between the target state X _R the designer wishes to calculate the manipulated variable tau _M by performing the state feedback.

When the control device as described above is actually configured, it is generally realized by using a digital signal processor (DSP) in order to obtain the advantages of the digital system. In this case, as shown in the figure of the present application. It is necessary to replace a continuous time system with a discrete time system.
Along with this discretization, there arises a problem that the control system becomes unstable due to calculation delay (= 1 sampling period delay) due to control processing. Therefore, the calculation delay of the manipulated variable is regarded as one state, and this is added to the state feedback, so that the calculation delay can be compensated and the stability of the control system can be maintained.
In the softly coupled two-inertia system, the disturbance torque applied to the driven shaft side by the disturbance estimator 5 can be estimated. Therefore, the estimation accuracy of the state estimator 7 is improved, and robust and stable feedback control is possible.
Further, since a disturbance torque detector and each state detector other than the observation output are not required, the configuration of the control system is simplified and the cost can be reduced. Moreover, by providing a disturbance memory, it is possible to store a disturbance situation estimated in the past.
Therefore, in the periodic disturbance, the disturbance can be removed up to a higher frequency band than in the case where there is no disturbance memory 6 due to the feedback loop. Furthermore, the state feedback controller 8 can realize high-precision rotation control by high gain control even for a controlled object having torsion.
By obtaining the estimated disturbance value 5, it is possible to accurately estimate each internal state of the controlled object. In addition, by storing the disturbance situation estimated in the past and obtaining the current disturbance estimated value from the stored contents, the responsiveness to the periodic disturbance can be enhanced.
All the internal states of the control object necessary for torsional vibration system control are prepared, and state feedback control is possible. In addition, state feedback control for torsional vibration system control is realized, and optimal control for the model becomes possible.
It is possible to eliminate instability of the control system due to calculation delay (= 1 sampling period delay) of the control processor. In addition, since a control object having a twist can be controlled beyond the resonance frequency, high gain control is realized, and highly accurate rotation control is possible.
Since the design method is unified, control design becomes easy. Compared with the pole placement method, the Kalman filter is easy to design because it is not necessary for the designer to consider poles and zeros, and only gives noise covariance data.
Compared with the pole placement method, the optimal regulator does not require the designer to consider pole zeros, and the design is easy because only the weighting factor is set.

It is a block diagram which shows the whole block diagram of the state feedback control system of the torsional vibration system control apparatus by this invention. FIG. 2 is a detailed diagram showing a disturbance estimator in FIG. 1. FIG. 2 is a detailed diagram illustrating the state estimator of FIG. 1. FIG. 2 is a detailed view showing the state feedback controller of FIG. 1. It is the schematic which shows the physical model of the two inertia system couple | bonded softly used as control object. It is the schematic which shows the structure which ignores the twist of an elastic body element generally and treats it as a rigid body.

Explanation of symbols

C Control unit 1 Motor 1 (drive side inertial body)
1a Torsion shaft (elastic element)
2 Inertial body (driven inertial body)
3 Angle detector 4 Angle detector 5 Disturbance estimator 6 Disturbance memory 7 State estimator 8 State feedback controller

Claims (3)

  A motor to which the first angle detector is connected;
  An inertial body connected to the motor and an elastic body element and connected to a second angle detector;
  The inertial body using a motor operation amount value input to the motor, a first angle value detected by the first angle detector, and a second angle value detected by the second angle detector. A disturbance estimator for calculating a torque value applied to
  A disturbance memory for storing the torque value calculated by the disturbance estimator;
  Control including the motor and the inertial body using the torque value stored in the disturbance memory, the motor operation amount value input to the motor, the first angle value, and the second angle value A state estimator that estimates an angle value and an angular velocity value that are internal states of the object;
  Based on a deviation between an angle value and an angular velocity value, which are internal states of the control object estimated by the state estimator, and an angle value and an angular velocity value, which are control target values for the motor and the inertial body, are input to the motor. A controller for calculating a motor manipulated variable value,
  A torsional vibration system control device comprising:   The state estimator is a state estimator that estimates an angle value and an angular velocity value that are internal states of the control target using a torque value stored in the disturbance memory when the inertial body makes one rotation. The torsional vibration system control device according to claim 1.   2. The controller according to claim 1, wherein the controller controls an upper limit value of a frequency of a control target including the motor and the inertial body to a value higher than a resonance frequency value of the control target. The torsional vibration system control device described.

Images