A method for suppressing periodic load fluctuation of voice coil actuator based on equivalent resonance embedding

Abstract

A method for suppressing periodic load fluctuations of voice coil actuators based on equivalent resonance embedding can achieve sensitive tracking and accurate compensation of periodic disturbances by embedding equivalent resonance terms of reciprocating motion. No additional load parameters are required, and the online estimation method is used, which is more suitable for application scenarios with large load conditions such as fast tool servo, fatigue testing and end effector of robotic arm.

Description

Technical Field

[0001] The present invention relates to the technical field of voice coil actuators, and in particular to a method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding. Background Art

[0002] Since voice coil actuators can directly drag loads to achieve reciprocating positioning motion in straight or arc trajectories, they do not require transmission devices such as gearboxes and cam linkages in typical electromechanical systems. They can significantly reduce the dead zone and hysteresis effect of the moving part's motion trajectory caused by mechanical clearance, thereby improving the positioning accuracy of the equipment. On the other hand, as a direct power source, voice coil actuators eliminate the transmission link, which can effectively reduce the inertia of the moving part, reduce the mechanical time constant of the moving part, and improve the response speed of the equipment. Due to their advantages of high positioning accuracy and high response speed, voice coil actuators are widely used in mechanical hard disk reading and positioning, lens focusing, fast tool servo drive, robot arm end effector, fatigue testing equipment and other application scenarios that require both positioning response speed and positioning accuracy. However, the operation mode of voice coil actuators directly driving loads brings challenges in control system design while bringing the above advantages. Specifically, the direct drive mode that eliminates the transmission link allows load fluctuations to act directly on the voice coil. Due to the small inertia of the voice coil, the effect of load fluctuations on positioning accuracy will be more significant. In addition, in actual operation, voice coil actuators are mostly used to achieve reciprocating motion trajectory following, and the design requirements for the position loop bandwidth are much higher than the bandwidth requirements of general motion servo. In a single cycle during the reciprocating motion process, the motion velocity curve of the voice coil actuator dragging the load will have at least two zero crossings. In this process, there is a phenomenon that the nonlinear load component represented by friction has a curve shock change, which will further increase the challenges faced by the design of the voice coil actuator controller. In the known public literature, researchers mainly proposed the following methods around the position servo problem of voice coil actuators: 1) Starting from the uncertainty of system modeling, clarify the upper and lower bounds of the overall system dynamic uncertainty of the electromechanical system through experiments or theoretical analysis, design a robust controller, and achieve the stabilization of the closed-loop servo system in the working frequency band. Although this type of method can ensure the stability of the closed-loop system, it sacrifices the positioning accuracy of the system to a certain extent. 2) Starting from the known structural nonlinear dynamics of the system, the positioning accuracy of the closed-loop system is improved by inverse model compensation of the known nonlinear dynamics. This type of method has a good inhibitory effect on the structural nonlinear dynamics of the system itself, but cannot provide inhibitory effect on friction loads with time-varying characteristics. 3) Starting from the description and compensation of generalized disturbances, the part outside the linear nominal dynamics of the system is defined as disturbance and compensated offline. This type of method can achieve good results in application scenarios such as mechanical hard disk reading positioning and lens focusing where the load dynamic characteristics are relatively stable, but it cannot cope with scenarios such as robot end effectors and fatigue testing equipment. Summary of the invention

[0003] In order to overcome the deficiencies of the above technologies, the present invention provides a method for overcoming the deficiencies existing in the application of voice coil actuator positioning control. Considering that voice coil actuators are mostly used for reciprocating motion trajectory positioning and external interference is mostly presented in a periodic characteristic, a voice coil actuator control method based on equivalent resonance embedding is proposed to solve the voice coil actuator positioning problem caused by periodic load fluctuations.

[0004] The technical solution adopted by the present invention to overcome the technical problems is:

[0005] A method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding comprises the following steps:

[0006] a) Establish the nominal transfer function model G of the voice coil actuator n (s);

[0007] b) Establish the k-th order representation R in the s-domain of the equivalent resonant term used to describe the periodic load disturbance k (s);

[0008] c) Represent the equivalent resonant term in the s domain R k (s) is combined into lumped equivalent resonance terms according to the resonance order in the s domain. h (s);

[0009] d) Build the s-domain representation of the Butterworth low-pass filter Q but (s);

[0010] e) Establish the s-domain representation Q of the periodic disturbance estimation filter br (s);

[0011] f) Establish the s-domain representation U(s) of the controller;

[0012] g) Convert the controller's s-domain representation U(s) to its z-domain representation U(z)

[0013] h) The discretized controller U(z) is deployed in the digital controller of the voice coil actuator.

[0014] Furthermore, in step a), the formula Establish the nominal transfer function model G of the voice coil actuator n (s), where K sis the force constant of the voice coil actuator, L is the voice coil inductance, M is the total mass of the voice coil, winding frame and spindle of the voice coil actuator, C is the damping coefficient between the winding frame and the magnetic cylinder of the voice coil actuator, R is the voice coil resistance, s is the s domain complex variable, K is the elastic coefficient of the internal positioning leaf spring of the voice coil actuator, and the voltage values ​​of the two ends of the voice coil of the voice coil actuator are input into the nominal transfer function model G n In (s), the output is the displacement value of the center point of the voice coil relative to the magnetic cylinder of the voice coil actuator.

[0015] Preferably, the force constant K of the voice coil actuator is s is the ratio of the thrust output of the voice coil actuator to the voice coil current. Further, in step b), the formula The k-th order representation R of the equivalent resonance term in the s domain is calculated. k (s), where s is a complex variable in the s domain, a k and b k are all adjustable parameters, k is the harmonic order, ω k is the reciprocating angular frequency of the voice coil actuator.

[0016] Furthermore, in step c), by formula The s-domain representation of the lumped equivalent resonance term is calculated as R h (s), where K h is the adjustable resonance order.

[0017] Furthermore, in step d), by formula The s-domain representation Q of the Butterworth low-pass filter is calculated but (s), where ω c is the cutoff angular frequency, s is the s-domain complex variable, N is the order of the Butterworth low-pass filter, and the formula k=0,1,...,N-1Calculate the full complex plane poles s of the Butterworth low-pass filter k , the full complex plane pole s k The left and right parts are equally divided to form the poles of the left and right half complex planes, s lk is the lkth pole in the left half complex plane, where e is a natural constant and j is the imaginary unit of the s domain.

[0018] Furthermore, in step e), by formula The s-domain representation Q of the periodic disturbance estimation filter is calculated br (s).

[0019] Furthermore, in step f), by formula Establish the s-domain representation of the controller U(s), where K p is an adjustable scale parameter, Y r(s) is the s-domain representation of the expected value of the position signal of the voice coil center point relative to the magnetic cylinder, Y(s) is the s-domain representation of the actual feedback signal of the position of the voice coil center point relative to the magnetic cylinder, K i is the adjustable integral parameter, K dt is the pre-adjustable proportional term for periodic disturbance compensation, and s is a complex variable in s domain.

[0020] Further, step g) comprises the following steps:

[0021] g-1) by formula

[0022]

[0023] Establish the nominal transfer function model G of the voice coil actuator n (z), where T is the discretized sampling time and z is a z-domain complex variable;

[0024] g-2) by formula

[0025]

[0026] The z-domain representation Q of the periodic disturbance estimation filter is calculated br (z);

[0027] g-3) Through the formula Establish the z-domain representation of the controller U(z), where Y r (z) is the z-domain representation of the expected value of the position signal of the center point of the voice coil relative to the magnetic cylinder, and Y(z) is the z-domain representation of the actual feedback signal of the position of the center point of the voice coil relative to the magnetic cylinder. The beneficial effect of the present invention is that by embedding the equivalent resonance term of the reciprocating motion, sensitive tracking and precise compensation of periodic disturbances can be achieved. There is no need to obtain additional load parameters, and the online estimation method is adopted, which is more suitable for application scenarios with large changes in load conditions such as fast tool servo, fatigue testing and robot end effector. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a structural diagram of the control system of the present invention;

[0029] Figure 2 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0030] The following is combined with Figure 1 , Attachment Figure 2 The present invention is further described.

[0031] A method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding comprises the following steps:

[0032] a) Establish the nominal transfer function model G of the voice coil actuator n (s);

[0033] b) Establish the k-th order representation R in the s-domain of the equivalent resonant term used to describe the periodic load disturbance k (s);

[0034] c) Represent the equivalent resonant term in the s domain R k (s) is combined into lumped equivalent resonance terms according to the resonance order in the s domain. h (s);

[0035] d) Build the s-domain representation of the Butterworth low-pass filter Q but (s);

[0036] e) Establish the s-domain representation Q of the periodic disturbance estimation filter br (s);

[0037] f) Establish the s-domain representation U(s) of the controller;

[0038] g) converting the controller's s-domain representation U(s) into a z-domain representation U(z);

[0039] h) The discretized controller U(z) is deployed in the digital controller of the voice coil actuator.

[0040] As attached Figure 1 As shown, the hardware part of the voice coil actuator is a well-known technology, which mainly includes a position sensor, a digital controller, a digital power amplifier, and a voice coil actuator. The position sensor uses a linear orthogonal encoder, an eddy current sensor, or a LVDT. The digital controller can use a digital signal processor or an MCU chip. The main circuit topology of the digital power amplifier is a two-phase full bridge, and the switch tube uses an IGBT or PowerMOSFET. The signal circuit from the digital controller to the digital power amplifier uses an optical coupling isolation circuit.

[0041] As attached Figure 2 As shown, unlike the general voice coil actuator control method that mostly draws on the traditional servo motor control framework, the present invention specifically proposes a method for suppressing periodic disturbances for the application scenario of multi-drive reciprocating motion of voice coil actuators. By embedding the equivalent resonance term of reciprocating motion, sensitive tracking and precise compensation of periodic disturbances can be achieved. There is no need to obtain additional load parameters, and the online estimation method is used, which is more suitable for application scenarios with large load conditions such as fast tool servo, fatigue testing and robotic arm end effector.

[0042] Embodiment 1:

[0043] In step a), the formula Establish the nominal transfer function model G of the voice coil actuator n (s), where K s is the force constant of the voice coil actuator, L is the voice coil inductance, M is the total mass of the voice coil, winding frame and spindle of the voice coil actuator, C is the damping coefficient between the winding frame and the magnetic cylinder of the voice coil actuator, R is the voice coil resistance, s is the s domain complex variable, K is the elastic coefficient of the internal positioning leaf spring of the voice coil actuator, and the voltage values ​​of the two ends of the voice coil of the voice coil actuator are input into the nominal transfer function model G n In (s), the output is the displacement value of the center point of the voice coil relative to the magnetic cylinder of the voice coil actuator.

[0044] Embodiment 2:

[0045] Force constant K of voice coil actuator s It is the ratio of the thrust output of the voice coil actuator to the voice coil current.

[0046] Embodiment 3:

[0047] In step b), the formula The k-th order representation R of the equivalent resonance term in the s domain is calculated. k (s), where s is a complex variable in the s domain, a k and b k are all adjustable parameters, k is the harmonic order, ω k is the reciprocating angular frequency of the voice coil actuator.

[0048] Embodiment 4:

[0049] In step c), the formula The s-domain representation of the lumped equivalent resonance term is calculated as R h (s), where K h is the adjustable resonance order.

[0050] Embodiment 5:

[0051] In step d), the formula The s-domain representation Q of the Butterworth low-pass filter is calculated but (s), where ω c is the cutoff angular frequency, s is the s-domain complex variable, N is the order of the Butterworth low-pass filter, and the formula k=0,1,...,N-1Calculate the full complex plane poles s of the Butterworth low-pass filter k , the full complex plane pole s k The left and right parts are equally divided to form the poles of the left and right half complex planes, s lk is the lkth pole in the left half complex plane, where e is a natural constant and j is the imaginary unit of the s domain.

[0052] Embodiment 6:

[0053] In step e), the formula The s-domain representation Q of the periodic disturbance estimation filter is calculated br (s).

[0054] Embodiment 7:

[0055] In step f), the formula Establish the s-domain representation of the controller U(s), where K p is an adjustable scale parameter, Y r (s) is the s-domain representation of the expected value of the position signal of the voice coil center point relative to the magnetic cylinder, Y(s) is the s-domain representation of the actual feedback signal of the position of the voice coil center point relative to the magnetic cylinder, K i is the adjustable integral parameter, K dt is the pre-adjustable proportional term for periodic disturbance compensation, and s is a complex variable in s domain.

[0056] Embodiment 8:

[0057] Step g) comprises the following steps:

[0058] g-1) by formula

[0059]

[0060] Establish the nominal transfer function model G of the voice coil actuator n (z), where T is the discretized sampling time and z is a complex variable in the z domain.

[0061] g-2) by formula

[0062]

[0063] The z-domain representation Q of the periodic disturbance estimation filter is calculated br (z);

[0064] g-3) Through the formula Establish the z-domain representation of the controller U(z), where Y r(z) is the z-domain representation of the expected value of the position signal of the center point of the voice coil relative to the magnetic cylinder, and Y(z) is the z-domain representation of the actual feedback signal of the position of the center point of the voice coil relative to the magnetic cylinder. Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding, characterized in that: The steps include: a) Establish the nominal transfer function model G of the voice coil actuator n (s); b) Establish the k-th order representation R in the s-domain of the equivalent resonant term used to describe the periodic load disturbance k (s); c) Represent the equivalent resonant term in the s domain R k (s) is combined into lumped equivalent resonance terms according to the resonance order in the s domain. h (s); d) Build the s-domain representation of the Butterworth low-pass filter Q but (s); e) Establish the s-domain representation Q of the periodic disturbance estimation filter br (s); f) Establish the s-domain representation U(s) of the controller; g) converting the controller's s-domain representation U(s) into a z-domain representation U(z); h) deploying the discretized controller U(z) in the digital controller of the voice coil actuator; In step a), the formula Establish the nominal transfer function model G of the voice coil actuator n (s), where Ks is the force constant of the voice coil actuator, L is the voice coil inductance, M is the total mass of the voice coil, winding frame and main shaft of the voice coil actuator, C is the damping coefficient between the winding frame and the magnetic cylinder of the voice coil actuator, R is the voice coil resistance, s is an s-domain complex variable, K is the elastic coefficient of the internal positioning leaf spring of the voice coil actuator, and the voltage values ​​of the two ends of the voice coil of the voice coil actuator are input into the nominal transfer function model G n (s), the output is the displacement value of the center point of the voice coil relative to the magnetic cylinder of the voice coil actuator; In step b), the formula The k-th order representation R of the equivalent resonance term in the s domain is calculated. k (s), where s is a complex variable in the s domain, a k and b k are all adjustable parameters, k is the harmonic order, ω k is the reciprocating angular frequency of the voice coil actuator.

2. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 1 is characterized in that: Force constant K of voice coil actuator s It is the ratio of the thrust output of the voice coil actuator to the voice coil current.

3. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 1, characterized in that: In step c), the formula The s-domain representation of the lumped equivalent resonance term is calculated as R h (s), where Kh is the adjustable resonance order.

4. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 3 is characterized in that: In step d), the formula The s-domain representation Q of the Butterworth low-pass filter is calculated but (s), where ω c is the cutoff angular frequency, s is the s-domain complex variable, N is the order of the Butterworth low-pass filter, and the formula Calculate the full complex plane poles s of the Butterworth low-pass filter k , the full complex plane pole s k The left and right parts are equally divided to form the poles of the left and right half complex planes, s lk is the lkth pole in the left half complex plane, where e is a natural constant and j is the imaginary unit of the s domain.

5. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 4 is characterized in that: In step e), the formula The s-domain representation Q of the periodic disturbance estimation filter is calculated br (s).

6. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 1, characterized in that: In step f), the formula Establish the s-domain representation of the controller U(s), where K p is an adjustable scale parameter, Y r (s) is the s-domain representation of the expected value of the position signal of the voice coil center point relative to the magnetic cylinder, Y(s) is the s-domain representation of the actual feedback signal of the position of the voice coil center point relative to the magnetic cylinder, K i is the adjustable integral parameter, K dt is the pre-adjustable proportional term for periodic disturbance compensation, and s is a complex variable in s domain.

7. The method for suppressing periodic load fluctuation of a voice coil actuator based on equivalent resonance embedding according to claim 6 is characterized in that: Step g) comprises the following steps: g-1) by formula Establish the nominal transfer function model G of the voice coil actuator n (z), where T is the discretized sampling time and z is a z-domain complex variable; g-2) by formula The z-domain representation Q of the periodic disturbance estimation filter is calculated br (z); g-3) Through the formula Establish the z-domain representation of the controller U(z), where Y r (z) is the z-domain representation of the expected value of the position signal of the center point of the voice coil relative to the magnetic cylinder, and Y(z) is the z-domain representation of the actual feedback signal of the position of the center point of the voice coil relative to the magnetic cylinder.

Images