CN107992110B - Magnetic suspension control moment gyro frame angular rate servo system based on harmonic reducer - Google Patents

Abstract

The invention proposes a magnetic levitation control torque gyro frame angular rate servo system based on a harmonic reducer, including a controller, a power amplifier, a torque motor, a linear Hall sensor at the motor end, a harmonic reducer, a load, a load-side rotary transformer, Cascading expansion state observers. Among them, the control output of the controller and the angular position of the load end are used as the input signals of the cascade expansion state observer, and additional control information is generated by the cascade expansion state observer to compensate the controller; the torque motor end is installed with a linear Hall sensor and passed through the expansion state observer. Solve to obtain accurate torque motor rotor position information, torque motor is fixedly connected to the input end of the harmonic reducer, the output torque of the harmonic reducer acts on the load, and a resolver is installed at the load end to measure the angular position of the load end. The magnetic suspension control torque gyro frame servo system proposed by the present invention can realize high-precision frame angular rate servo control.

Description

A Magnetic Levitation Control Torque Gyro Frame Angular Rate Servo Based on Harmonic Reducer system

technical field

The invention belongs to the field of magnetic suspension control torque gyro angular rate servo control based on harmonic reducers, and in particular relates to a magnetic suspension control torque gyro high-precision frame angular rate servo system using a composite control algorithm based on cascaded expansion state observers, which is beneficial to Improve the angular rate servo control accuracy of the overall system.

Background technique

For the magnetic suspension control torque gyro frame servo system, if the direct drive method is adopted, the volume and weight of the torque motor will also increase accordingly. In order to reduce the volume and weight of the frame system, a transmission mechanism is generally used to amplify the motor torque. The harmonic reducer has the advantages of large transmission ratio, high precision, small size and light weight, and is one of the best choices for controlling the torque gyro frame transmission mechanism. However, the nonlinear characteristics of the harmonic reducer such as nonlinear torsional stiffness, non-linear Linear friction, backlash, etc. seriously affect the angular rate accuracy of the frame system.

In order to solve the problem that the non-linear characteristics of the harmonic reducer lead to the reduction of the angular rate accuracy of the frame system, in the Chinese patent a magnetic suspension control torque gyro frame system based on the harmonic reducer (patent number: ZL201310435526.8), the output of the torque motor The shaft is fixedly connected with the input end of the harmonic reducer. The harmonic reducer is used as a torque amplifying device, and the output torque acts on the load. The type and installation position of the sensor usually adopt the "photoelectric code disk installed on the torque motor end and the load end, respectively." The structure of measuring the angular position of the motor end and the angular rate of the load end", the control algorithm based on the torsional stiffness hysteresis model and compensation of the harmonic reducer is used to suppress the inherent hysteresis characteristics of the harmonic reducer in the magnetic suspension control torque gyro frame servo system Influence on the accuracy of the system, but using the photoelectric encoder as a sensor will reduce the environmental adaptability and service life of the entire frame system, and the parameters of the hysteresis model in the control algorithm are generally difficult to accurately identify; in order to solve the nonlinear characteristics of the harmonic reducer The problem that leads to the reduction of the angular rate accuracy of the frame system, the type and installation position of the existing sensors usually adopt the structure of "only install the resolver at the load end to measure the angular position of the load end to design the controller algorithm", and the expansion state is adopted in the control algorithm. The observer combines the controller to estimate and compensate the nonlinear transmission torque of the harmonic reducer, but the rotor position information of the torque motor cannot be obtained only by installing the resolver at the load end, and the control algorithm exists when the system order is high, the state observer can be expanded. The parameters are more difficult to adjust.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is: to overcome the problem that the output rate accuracy of the existing magnetic suspension control torque gyro angular rate servo system based on the harmonic reducer is not high, and to provide a structural scheme for improving the output angular rate accuracy of the frame system; control algorithm.

The technical solution adopted by the present invention to solve the above technical problems is: a magnetic suspension control torque gyro frame angular rate servo system based on a harmonic reducer, comprising a controller, a power amplifier, a torque motor, a linear Hall sensor, a rotor position calculation module, Harmonic reducer, magnetic levitation control torque gyro frame system load, resolver, angular position solution module, cascade expansion state observer; among them, the controller uses a composite control algorithm based on state feedback and disturbance compensation to generate control signals, The actual control current is output to the torque motor through the power amplifier. The linear Hall sensor installed at the torque motor end is used to measure the relevant information of the rotor position of the torque motor. The rotor position of the torque motor is obtained through the rotor position calculation module. The output shaft of the torque motor is decelerated by harmonics. The input end of the inverter is fixedly connected, and the harmonic reducer is used as a torque amplifying device. The output torque acts on the load of the magnetic suspension control torque gyro frame system. The resolver at the load end obtains the angular position of the load end through the angular position calculation module. The output of the controller control is used as the input information of the cascade expansion state observer. The cascade expansion state observer generates additional control variables, which are combined with the angular rate command of the magnetic suspension control torque gyro frame angular rate servo system and the torque motor rotor given by the attitude control computer. The position and the angular position of the load end are taken together as input to the controller.

Wherein, the described composite control algorithm steps are as follows:

Step 1: Build the System Mathematical Model

The state variables of the system dynamics model based on the harmonic reducer are taken as:

X=[x ₁ x ₂ x ₃ x ₄ ] ^T =[θ _l ω _l Δθ ω _m ] ^T , the control input is taken as u=T _m , and the state space equation is:

Among them, θ _l is the angular position of the load end, ω _l is the angular velocity of the load end obtained by differentiating the angular position of the load end, ω _m is the angular velocity of the motor end obtained by differentiating the rotor position of the torque motor θ _m , and n is the harmonic reducer Δθ=θ _m /n-θ _l is the torsion angle of the harmonic reducer; T _m is the electromagnetic torque of the motor, T _f is the disturbance torque; J _m and J _l are the moment of inertia of the motor and the load, respectively; B _m and B _l are the damping coefficients of the motor and the load, respectively; K _h is the linear time-invariant part of the torsional stiffness of the harmonic reducer, and ΔK _h is the nonlinear time-varying part of the torsional stiffness of the harmonic reducer.

The coordinate transformation of the mathematical model of the system is as follows:

The state space expression is:

in,

Step 2: Cascade Extended State Observer Design

Define the state variables of the cascaded dilated state observers as z = [z ₁ , z ₂ , z ₃ , z ₄ , z ₅ , z ₆ , z ₇ , z ₈ ] ^T , where z ₁ estimates v ₁ , z ₂ estimates v ₂ , z ₄ estimates v ₃ , z ₆ estimates v ₄ , z ₈ estimates f', and z ₃ , z ₅ , and z ₇ are intermediate variables of the cascaded extended state observer.

The state equation of the cascade expansion state observer is:

in,

补偿到控制器(1)中的控制律中。β ₁ , β ₂ are two design parameters of the cascade expansion state observer. The cascade expansion state observer produces additional control variables through the above state equation

Compensated into the control law in the controller (1).

Using the cascaded form of four second-order extended state observers with the same parameters, the observer needs to configure only two parameters, namely: β ₁ and β ₂ , which can solve the problem that the observer parameters are difficult to configure.

Step 3: Composite Control Algorithm Design

The following control law is adopted in the controller (1):

Among them, ω _lref is the angular rate command, θ _lref is the reference rotor position information obtained by integrating the angular rate command, v ₁ =θ _l is the angular position of the load end, v ₂ =ω _l is the angular velocity of the load end, z ₄ , z ₆ , z ₈ are the additional control variables generated by the cascade expansion state observer; k ₁ , k ₂ , k ₃ , and k ₄ are the four design parameters of the controller, b is the frame system parameter, and u is the control output of the controller .

Step 4: Parameter design of composite control algorithm

There are six parameters that need to be configured in the whole control algorithm, namely: controller parameters k ₁ , k ₂ , k ₃ , k ₄ and two parameters β ₁ and β ₂ of the cascade expansion state observer. Among them, the controller parameters k ₁ , k ₂ , k ₃ , and k ₄ are performed according to the traditional pole configuration, while the two parameters β ₁ and β ₂ of the cascaded extended state observer are designed according to the disturbance of the specific system:

Among them, ζ is the damping ratio of the system, and ω _f is the rotational frequency of the magnetic suspension control torque gyro magnetic suspension rotor.

The basic principle of the present invention is: the present invention installs a small-sized and light-weight linear Hall sensor on the torque motor end, obtains the accurate torque motor rotor position through the rotor position calculation module, the torque motor is fixedly connected to the harmonic reducer, and the harmonic deceleration The output torque of the sensor acts on the load, a resolver is installed at the load end, and the angular position of the load end is measured by the angular position calculation module; a composite control algorithm based on the cascaded expansion state observer is adopted, and the cascaded expansion state observer is used for the system state and disturbance. The estimation is performed and additional control information is generated to compensate the controller, thereby suppressing the system disturbance and realizing the high-precision frame angular rate output.

The advantages of the present invention compared with the prior art are:

1. Compared with the existing magnetic levitation control torque gyro frame servo system, the present invention adopts a small and lightweight linear Hall sensor to measure the angular position of the motor end, and adopts a highly reliable resolver to measure the angular position of the load end, which can not only overcome the The problem that the rotor position information of the torque motor and the angular position information of the load end cannot be accurately measured at the same time in the existing system can also improve the environmental adaptability of the entire frame servo system.

2. Using a composite control algorithm based on cascaded extended state observers, four second-order extended state observers with the same parameters are cascaded. The observer needs to configure only two parameters, namely: β ₁ and β ₂ , It can well solve the problem that there are many parameters that are difficult to adjust in the existing controller design methods such as establishing a compensation model and combining the controller with an expanded state observer. By designing the control parameters, the overall anti-interference ability of the system is improved, and the output angular rate accuracy of the system is improved.

Description of drawings

Fig. 1 is the frame angular rate servo control system control diagram;

Figure 2 is a schematic diagram of a frame system based on a harmonic reducer;

Figure 3 is the output signal of the linear Hall sensor;

Fig. 4 is the rotor position calculation module based on linear Hall sensor;

FIG. 5 is a structural diagram of the cascade expansion state observer.

The meanings of the reference symbols in the figure are: 1 is the controller, 2 is the power amplifier, 3 is the torque motor, 4 is the linear Hall sensor, 5 is the rotor position calculation module, 6 is the harmonic reducer, and 7 is the magnetic suspension control torque gyro The frame system load, 8 is the resolver, 9 is the angular position solution module, and 10 is the cascade expansion state observer.

Detailed ways

The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

As shown in Figure 1, a magnetic levitation control torque gyro frame angular rate servo system based on harmonic reducer includes a controller 1, a power amplifier 2, a torque motor 3, a linear Hall sensor 4, a rotor position calculation module 5, a harmonic Wave reducer 6, magnetic levitation control torque gyro frame system load 7, resolver 8, angular position calculation module 9, cascade expansion state observer 10; wherein, the controller 1 adopts a composite control based on the combination of state feedback and disturbance compensation The algorithm generates a control signal, and outputs the actual control current to the torque motor 3 through the power amplifier 2. The linear Hall sensor 4 installed at the torque motor end is used to measure the information related to the rotor position of the torque motor, and the rotor position of the torque motor is obtained through the rotor position calculation module 5. The output shaft of the torque motor 3 is fixedly connected with the input end of the harmonic reducer 6. The harmonic reducer 6 is used as a torque amplifying device, and the output torque acts on the magnetic suspension control torque gyro frame system load 7. The resolver 8 at the load end is solved by the angular position. The calculation module 9 obtains the angular position of the load end, and the angular position of the load end together with the control output of the controller is used as the input information of the cascade expansion state observer 10, and the cascade expansion state observer 10 generates an additional control quantity, which is the same as that given by the attitude control computer. The angular rate command of the magnetic suspension control torque gyro frame angular rate servo system, the rotor position of the torque motor and the angular position of the load end are used as the input information of the controller 1 together.

FIG. 2 is a schematic diagram of the mechanical connection of the torque motor 3 , the linear Hall sensor 4 , the harmonic reducer 6 , the magnetic suspension control torque gyro frame system load 7 , and the resolver 8 in FIG. 1 . The linear Hall sensor is installed on the torque motor end, the torque motor is fixedly connected to the input end of the harmonic reducer, the harmonic reducer is connected to the load of the magnetic levitation control torque gyro frame system, and a resolver is installed at the load end.

The acquisition process of the rotor position of the torque motor in Figure 1 is as follows:

Step 1: The linear Hall sensor obtains the signal related to the rotor position of the torque motor. Two linear hall sensors (linear hall sensor A and linear hall sensor B) with an electrical angle difference of 90° are installed on one end of the motor. When the motor rotates, the output signals of the two linear hall sensors are theoretically two The sine signal V _sin and the cosine signal V _cos of the rotor position electrical angle different from each other by 90° are shown in FIG. 3 .

如图4所示，线性霍尔传感器A和线性霍尔传感器B的信号经过信号调理电路，经过AD转换(数字/模拟转换)之后经计算得到当前的力矩电机转子位置

(电角度)。当前力矩电机转子位置

(电角度)计算公式可表示为：Step 2: Obtain the rotor position of the torque motor through the torque motor rotor position calculation module 5

As shown in Figure 4, the signals of the linear Hall sensor A and the linear Hall sensor B pass through the signal conditioning circuit, and after AD conversion (digital/analog conversion), the current torque motor rotor position is obtained by calculation

(electrical angle). Current torque motor rotor position

The (electrical angle) calculation formula can be expressed as:

The steps of the composite control algorithm are as follows:

Step 1: Build the System Mathematical Model

The state variables of the system dynamics model based on the harmonic reducer are taken as:

X=[x ₁ x ₂ x ₃ x ₄ ] ^T =[θ _l ω _l Δθ ω _m ] ^T ,

The control input is taken as u=T _m , and the state space equation is:

是谐波减速器扭转刚度中的线性时不变部分，ΔK_h是谐波减速器扭转刚度中的非线性时变部分。Among them, θ _l is the angular position of the load end, ω _l is the angular velocity of the load end obtained by differentiating the angular position of the load end, ω _m is the angular velocity of the motor end obtained by differentiating the rotor position of the torque motor θ _m , and n is the harmonic reducer Δθ=θ _m /n-θ _l is the torsion angle of the harmonic reducer; T _m is the electromagnetic torque of the motor, T _f is the disturbance torque; J _m and J _l are the moment of inertia of the motor and the load, respectively; B _m and B _l are the damping coefficients of the motor and the load, respectively;

is the linear time-invariant part of the torsional stiffness of the harmonic reducer, and ΔK _h is the nonlinear time-varying part of the torsional stiffness of the harmonic reducer.

The coordinate transformation of the mathematical model of the system is as follows:

The state space expression is:

in,

Step 2: Cascade Extended State Observer Design

Define the state variables of the cascaded dilated state observers as z = [z ₁ , z ₂ , z ₃ , z ₄ , z ₅ , z ₆ , z ₇ , z ₈ ] ^T , where z ₁ estimates v ₁ , z ₂ estimates v ₂ , z ₄ estimates v ₃ , z ₆ estimates v ₄ , z ₈ estimates f', and z ₃ , z ₅ , and z ₇ are intermediate variables of the cascaded extended state observer.

The state equation of the cascade expansion state observer is:

in,

补偿到控制器(1)中的控制律中。β ₁ , β ₂ are two design parameters of the cascade expansion state observer. The structure of the cascade expansion state observer is shown in Figure 5. The cascade expansion state observer generates additional control variables through the above state equation

Compensated into the control law in the controller (1).

Using the cascaded form of four second-order extended state observers with the same parameters, the observer needs to configure only two parameters, namely: β ₁ and β ₂ , which can solve the problem that the observer parameters are difficult to configure.

Step 3: Composite Control Algorithm Design

The following control law is adopted in the controller (1):

Among them, ω _lref is the angular rate command, θ _lref is the reference rotor position information obtained by integrating the angular rate command, v ₁ =θ _l is the angular position of the load end, v ₂ =ω _l is the angular velocity of the load end, z ₄ , z ₆ , z ₈ are the additional control variables generated by the cascade expansion state observer; k ₁ , k ₂ , k ₃ , and k ₄ are the four design parameters of the controller, b is the frame system parameter, and u is the control output of the controller .

Step 4: Parameter design of composite control algorithm

There are six parameters that need to be configured in the whole control algorithm, namely: controller parameters k ₁ , k ₂ , k ₃ , k ₄ and two parameters β ₁ and β ₂ of the cascade expansion state observer. Among them, the controller parameters k ₁ , k ₂ , k ₃ , and k ₄ are performed according to the traditional pole configuration, while the two parameters β ₁ and β ₂ of the cascaded extended state observer are designed according to the disturbance of the specific system:

Among them, ζ is the damping ratio of the system, and ω _f is the rotational frequency of the magnetic suspension control torque gyro magnetic suspension rotor.

By adopting the above control algorithm, the rate servo precision of the frame system can be improved.

Taking the angular rate servo system of magnetic levitation control torque gyro frame based on harmonic reducer with angular momentum of 200Nms as an example, the angular rate bandwidth is 10Hz. The system parameters are shown in Table 1.

Table 1 System parameters

The controller parameters k ₁ , k ₂ , k ₃ , and k ₄ are configured as:

k ₁ =1.56×10 ⁷ , k ₂ =9.92×10 ⁵ , k ₃ =2.37×10 ⁴ , k ₄ =251.33.

The rotation frequency of the magnetic suspension control torque gyro magnetic suspension rotor is 200Hz, that is, ω _f =2×π×200(rad/s), and the system damping ratio is ζ=0.707. Therefore, the two parameters β ₁ , β of the cascade expansion state observer ₂ is configured as:

It has been verified by simulation that when the angular rate command is 5°/s, the system output angular rate rate fluctuation in the present invention is only 0.01°/s, and the Chinese patent a dual-frame control torque gyro frame servo based on harmonic reducer. Compared with the system (patent number: ZL201310435526.8), the system output angular rate accuracy in the present invention is improved by 11.3%.

Claims (1)

1. a magnetic levitation control torque gyro frame angular rate servo system based on harmonic reducer, is characterized in that: comprise controller (1), power amplifier (2), torque motor (3), linear Hall sensor (4), Rotor position calculation module (5), harmonic reducer (6), magnetic suspension control torque gyro frame system load (7), resolver (8), angular position calculation module (9), cascade expansion state observer ( 10); wherein, the controller (1) uses a composite control algorithm based on the combination of state feedback and disturbance compensation to generate a control signal, and outputs the actual control current to the torque motor (3) through the power amplifier (2), and the linear torque motor installed at the end of the torque motor is used. The sensor (4) is used to measure the information related to the rotor position of the torque motor, and the rotor position of the torque motor is obtained through the rotor position calculation module (5). The output shaft of the torque motor (3) is fixed to the input end of the harmonic reducer (6). The harmonic reducer (6) acts as a torque amplifying device, and the output torque acts on the magnetic suspension control torque gyro frame system load (7), and the resolver (8) at the load end obtains the angular position of the load end through the angular position calculation module (9). , the angular position of the load end and the control output of the controller are used as the input information of the cascade expansion state observer (10), and the cascade expansion state observer (10) generates an additional control amount, which is related to the magnetic suspension control torque gyro given by the attitude control computer. The angular rate command of the frame angular rate servo system, the rotor position of the torque motor and the angular position of the load end are taken together as the input information of the controller (1); The described composite control algorithm steps are as follows: Step 1: Build the System Mathematical Model The state variables of the system dynamics model based on the harmonic reducer are taken as: X=[x ₁ x ₂ x ₃ x ₄ ] ^T =[θ _l ω _l Δθ ω _m ] ^T , the control input is taken as u=T _m , and the state space equation is:

Among them, θ _l is the angular position of the load end, ω _l is the angular velocity of the load end obtained by differentiating the angular position of the load end, ω _m is the angular velocity of the motor end obtained by differentiating the rotor position of the torque motor θ _m , and n is the harmonic reducer Δθ=θ _m /n-θ _l is the torsion angle of the harmonic reducer; T _m is the electromagnetic torque of the motor, T _f is the disturbance torque; J _m and J _l are the moment of inertia of the motor and the load, respectively; B _m and B _l are the damping coefficients of the motor and the load, respectively;

is the linear time-invariant part of the torsional stiffness of the harmonic reducer, and ΔK _h is the nonlinear time-varying part of the torsional stiffness of the harmonic reducer; The coordinate transformation of the mathematical model of the system is as follows: v ₁ =x ₁ , v ₂ =x ₂ ,

The state space expression is:

in,

Step 2: Cascade Extended State Observer Design Define the state variables of the cascaded dilated state observers as z = [z ₁ , z ₂ , z ₃ , z ₄ , z ₅ , z ₆ , z ₇ , z ₈ ] ^T , where z ₁ estimates v ₁ , z ₂ estimates v ₂ , z ₄ estimates v ₃ , z ₆ estimates v ₄ , z ₈ estimates f', and z ₃ , z ₅ , and z ₇ are intermediate variables of the cascade expansion state observer; The state equation of the cascade expansion state observer is:

in,

β ₁ , β ₂ are the two design parameters of the cascade expansion state observer, and the cascade expansion state observer generates additional control variables through the above state equation

compensated into the control law in the controller (1); Using the cascaded form of four second-order extended state observers with the same parameters, the observer needs to be configured with only two parameters, namely: β ₁ and β ₂ , which can solve the problem that the observer parameters are difficult to configure; Step 3: Composite Control Algorithm Design The following control law is adopted in the controller (1):

Among them, ω _lref is the angular rate command, θ _lref is the reference rotor position information obtained by integrating the angular rate command, v ₁ =θ _l is the angular position of the load end, v ₂ =ω _l is the angular velocity of the load end, z ₄ , z ₆ , z ₈ are the additional control variables generated by the cascade expansion state observer; k ₁ , k ₂ , k ₃ , and k ₄ are the four design parameters of the controller, b is the frame system parameter, and u is the control output of the controller ; Step 4: Parameter design of composite control algorithm There are six parameters that need to be configured in the whole control algorithm, namely: controller parameters k ₁ , k ₂ , k ₃ , k ₄ and two parameters β ₁ , β ₂ of the cascade expansion state observer, among which, the controller parameter k ₁ , k ₂ , k ₃ , and k ₄ are performed according to the traditional pole configuration, while the two parameters β ₁ , β ₂ of the cascade extended state observer are designed according to the disturbance of the specific system:

Among them, ζ is the damping ratio of the system, and ω _f is the rotational frequency of the magnetic suspension control torque gyro magnetic suspension rotor.

Images