Model-free predictive control method for piezoelectric microactuators based on online data
Through the model-free prediction control method of online data, dynamic modal decomposition and secondary planning optimization, the problems of large computing volume and low accuracy in piezoelectric microdriver control are solved, and ultra-precision motion control is realized, suitable for multiple high-quality Accuracy application scenarios.
Patent Information
- Application Number
- CN202210988144.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-17
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-08-17
AI Technical Summary
In the prior art, the control of the piezoelectric microdriver has problems such as large calculation amount and poor prediction accuracy, which makes it difficult to achieve high-performance control in ultra-precision motion control.
Using a model-free prediction control method based on online data, the linear approximation dynamics system is solved through dynamic modal decomposition method, the model-free prediction controller is designed, and the control sequence is optimized through the secondary planning problem to achieve ultra-precision of the piezoelectric microdriver Motion control.
It reduces the computing volume of the controller and improves the prediction accuracy. It is suitable for embedded controllers with weak computing capabilities, and is suitable for piezoelectric microdrive control in semiconductor processing, manufacturing, aerospace and other fields.
Smart Images

Figure CN115236988B_ABST
Abstract
Description
Technical Field
[0001] The disclosed embodiments relate to the field of electrical technology, and in particular to a model-free predictive control method for a piezoelectric microactuator based on online data. Background Art
[0002] At present, with the emergence of micro / nanotechnology, people are trying to explore, design and manufacture on smaller and smaller physical scales. Nowadays, there is a demand for ultra-precision motion control in many fields such as semiconductor manufacturing, bioscience, aerospace, etc., which requires motion control of micron or even nanometer scale positioning. This ultra-precision motion usually requires the use of new smart materials-piezoelectric ceramics. Because piezoelectric microactuators have the advantages of infinite displacement resolution, high speed, high bandwidth, high efficiency and low heat generation. More and more ultra-precision micro / nano positioning platforms based on piezoelectric microactuators are put into practical application. However, due to the inherent creep and hysteresis nonlinearity of piezoelectric microactuators and the lack of understanding of their essential mechanisms, the control of piezoelectric microactuators has three major problems: difficult modeling, difficult prediction, and difficult control. In addition, its dynamic model inevitably has errors and uncertainties, which further increases the difficulty of designing high-performance controllers for piezoelectric microactuators. How to use advanced control technology to solve the above problems and achieve high control performance is a problem that researchers in this field need to solve.
[0003] In traditional industrial application scenarios, PID-based methods are mostly used for control. PID-based methods have a simple structure and do not require prior knowledge of the system model, but are often not high enough in precision control applications. In order to describe the nonlinearity of piezoelectric microactuators, researchers have proposed a variety of models such as the Preisach model, the Duhem model, and the Bouc–Wen model. Based on different piezoelectric ceramic dynamic models, researchers have designed a variety of advanced model-based control methods such as feedforward control, sliding mode control, adaptive sliding mode control, and model reference adaptive control. Although this type of model-based control method has achieved good results, there is still no unified model that can fully describe the dynamics of piezoelectric microactuators. And generally speaking, in order to obtain better modeling accuracy and higher control performance, the complexity of the model will be greatly increased, so an accurate model is usually accompanied by a large number of parameters that need to be identified. The identification and fine-tuning of a large number of parameters is time-consuming and labor-intensive. Therefore, the practicality of this type of model-based control method in actual engineering is greatly reduced.
[0004] It can be seen that there is an urgent need for a model-free predictive control method for piezoelectric microactuators based on online data with low computational complexity and accurate prediction. Summary of the invention
[0005] In view of this, an embodiment of the present disclosure provides a model-free predictive control method for a piezoelectric micro-actuator based on online data, which at least partially solves the problems of large amount of calculation and poor prediction accuracy in the prior art.
[0006] In a first aspect, an embodiment of the present disclosure provides a model-free predictive control method for a piezoelectric microactuator based on online data, comprising:
[0007] Step 1, giving the desired reference trajectory of the piezoelectric microactuator control system;
[0008] Step 2, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtain the linear approximation model corresponding to the piezoelectric micro-actuator, as well as the system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the linear approximation dynamic system;
[0009] Step 3: Design a model-free predictive controller based on online data according to the linear approximation model, and substitute the reference trajectory, system matrix A, control input matrix B, current state x(k), and weight matrix Q, W into the model-free predictive controller to obtain the optimal control sequence;
[0010] Step 4, using the first control variable of the optimal control sequence as the optimal control input for the next step to control the piezoelectric micro-actuator control system.
[0011] According to a specific implementation of the embodiment of the present disclosure, step 2 specifically includes:
[0012] Step 2.1, define the data structure of the sampling data;
[0013] Step 2.2, inputting an initial control signal I(t) to perform initial sampling on the piezoelectric microactuator;
[0014] Step 2.3, after the initial sampling is completed, the sampled data is used to form a linear approximation dynamic system of the decompressed electric micro-actuator;
[0015] Step 2.4, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtain the linear approximation model corresponding to the piezoelectric micro-actuator.
[0016] According to a specific implementation of the embodiment of the present disclosure, step 2.1 specifically includes:
[0017] Define multiple queues corresponding to the sampled data and the data update rules for each queue.
[0018] According to a specific implementation of the embodiment of the present disclosure, step 2.2 specifically includes:
[0019] Define the sampling interval and calculate the sampling period based on the preset sampling quantity;
[0020] Starting from inputting the initial control signal I(t), the system state variables of the piezoelectric micro-actuator and the values of the external control input at different times are sampled within the sampling period.
[0021] According to a specific implementation of the embodiment of the present disclosure, step 3 specifically includes:
[0022] Step 3.1, design a model-free predictive controller based on online data according to the linear approximation model and transform it into a constrained quadratic programming problem;
[0023] Step 3.2, calculate the parameters of the quadratic programming problem based on the reference trajectory, system matrix A, control input matrix B, current state x(k) and weight matrix Q, W;
[0024] Step 3.3, after substituting all parameters, solve the quadratic programming problem by the interior point method to obtain the optimal control sequence.
[0025] According to a specific implementation method of the embodiment of the present disclosure, the expression of the quadratic programming problem is:
[0026]
[0027]
[0028] Among them, the parameters H,f of the quadratic programming problem T The definition is as follows:
[0029] H=2(θQθ+W),f T =2E T Qθ
[0030] E=Ψx(k)-R k , Where A and B are the system matrix and control input matrix of the linear approximation dynamics system, p is the control time domain length of the model predictive control, x(k) is the system state at the current time k, R k is the reference trajectory sequence.
[0031] The piezoelectric micro-actuator model-free predictive control scheme based on online data in the embodiment of the present disclosure includes: step 1, giving the desired reference trajectory of the piezoelectric micro-actuator control system; step 2, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtaining the linear approximation model corresponding to the piezoelectric micro-actuator, as well as the system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the linear approximation dynamic system; step 3, designing a model-free predictive controller based on online data according to the linear approximation model, and substituting the reference trajectory, system matrix A, control input matrix B, current state x(k) and weight matrix Q, W into the model-free predictive controller to obtain the optimal control sequence; step 4, using the first control quantity of the optimal control sequence as the optimal control input for the next step to control the piezoelectric micro-actuator control system.
[0032] The beneficial effects of the embodiments of the present disclosure are as follows: through the scheme of the present disclosure, the I / O data of the system is collected, and based on the dynamic mode decomposition method with control items, an approximation solver is constructed to solve the linear approximation dynamic system, thereby realizing the discrete dynamic linearization of the unknown dynamic system. Then, based on the linear approximation model, a model predictive controller is designed to perform ultra-precision motion control on the piezoelectric micro-actuator. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0034] Figure 1 A schematic flow chart of a model-free predictive control method for a piezoelectric microactuator based on online data provided in an embodiment of the present disclosure;
[0035] Figure 2 An algorithm flow chart of a piezoelectric micro-actuator model-free predictive control method based on online data provided in an embodiment of the present disclosure;
[0036] Figure 3 A closed-loop control structure diagram of a piezoelectric micro-actuator involved in a piezoelectric micro-actuator model-free predictive control method based on online data provided in an embodiment of the present disclosure;
[0037] Figure 4 A schematic diagram of a setting control effect and control error provided by an embodiment of the present disclosure;
[0038] Figure 5 A trajectory tracking control effect and its control error are provided in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0039] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
[0040] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.
[0041] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.
[0042] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.
[0043] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.
[0044] The embodiments of the present disclosure provide a model-free predictive control method for a piezoelectric microactuator based on online data, which can be applied to the control process of a piezoelectric microactuator in scenarios such as semiconductor processing and manufacturing, ultra-precision measurement, aerospace, biomedicine, and nanoscience research.
[0045] See also Figure 1, is a flow chart of a piezoelectric micro-actuator model-free predictive control method based on online data provided by an embodiment of the present disclosure. Figure 1 As shown, the method mainly comprises the following steps:
[0046] Step 1, giving the desired reference trajectory of the piezoelectric microactuator control system;
[0047] When implementing it, the specific algorithm is as follows Figure 2 As shown in Figure 2, assuming that a piezoelectric micro-actuator control system has n state variables and m control inputs, the unknown piezoelectric micro-actuator dynamic model is: in represents the true value of each state variable of the piezoelectric micro-actuator control system at time t, represents the control input of the piezoelectric micro-actuator control system at time t. In addition, the reference trajectory of the piezoelectric micro-actuator control system is defined as:
[0048]
[0049] The reference trajectory represents the expected value of each state variable of the piezoelectric micro-actuator control system at time t, and the tracking error of the system is defined as:
[0050]
[0051] Step 2, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtain the linear approximation model corresponding to the piezoelectric micro-actuator, as well as the system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the linear approximation dynamic system;
[0052] Based on the above embodiment, step 2 specifically includes:
[0053] Step 2.1, define the data structure of the sampling data;
[0054] Step 2.2, inputting an initial control signal I(t) to perform initial sampling on the piezoelectric microactuator;
[0055] Step 2.3, after the initial sampling is completed, the sampled data is used to form a linear approximation dynamic system of the decompressed electric microactuator;
[0056] Step 2.4, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtain the linear approximation model corresponding to the piezoelectric micro-actuator.
[0057] Furthermore, the step 2.1 specifically includes:
[0058] Define multiple queues corresponding to the sampled data and the data update rules for each queue.
[0059] Furthermore, the step 2.2 specifically includes:
[0060] Define the sampling interval and calculate the sampling period based on the preset sampling quantity;
[0061] Starting from inputting the initial control signal I(t), the system state variables of the piezoelectric micro-actuator and the values of the external control input at different times are sampled within the sampling period.
[0062] In the specific implementation, the linear approximation dynamics of the piezoelectric micro-actuator is solved in real time, and this process is called dynamic linearization. In the present invention, we will use the dynamic mode decomposition method with control terms to solve the linear approximation dynamics. The steps are as follows:
[0063] S2.1 Define the data structure of the sampled data
[0064] In order to realize the dynamic linearization of the piezoelectric microactuator, in the actual algorithm, it is necessary to first define three queues, X, X', and Y. These three queues follow the first-in-first-out rule when updating the data elements therein: when a new data is sampled, the new sampled data can only enter from the head of the queue, and at the same time, an old data is taken out from the tail of the queue and deleted.
[0065] S2.2 Input the initial control signal I(t) for initial sampling
[0066] like Figure 3 As shown in the figure, before the control system starts running, since the input and output data of the system have not yet been obtained, it is necessary to give a small initial input signal and collect s groups of initial input and output sampling data. Assuming that the sampling interval of the system is Δt, the initial input signal must last at least for T = (s-1) Δt. If the initial signal is I(t), the external input of the closed-loop control system is recorded as:
[0067]
[0068] Let x(t) and u(t) represent the system state variables and external control input values of the piezoelectric micro-actuator at different times. After the initial sampling is completed, an initial queue storing the initial sampling data will be obtained as shown below:
[0069] X0=[x(0),x(Δt),...,x((s-1)Δt)],
[0070] X'0=[x(Δt),x(2Δt),…,x(sΔt)],
[0071] Y0=[u(0),u(Δt),…,u((s-1)Δt)]
[0072] These initial sampling data will be used to derive the initial linear approximation dynamics of the unknown system.
[0073] S2.3 Solving the linear approximation dynamics
[0074] Due to the dynamic system of the piezoelectric microactuator It is a nonlinear dynamic system, and it is generally believed that the dynamic system of the piezoelectric microactuator is a high-order continuously differentiable system. In traditional control methods, for such high-order continuously differentiable dynamic systems, the first-order Taylor expansion and other techniques can be used at the operating point to linearize them and approximate them as a linear system as follows:
[0075]
[0076] However, this linearization method requires the accurate dynamic model to be known in advance. This method cannot be used for linearization if the condition is unknown.
[0077] In order to realize the linearization of the system under the premise of unknown dynamic model, in the present invention, the linear approximation dynamics is solved based on the principle of dynamic mode decomposition method with control term. The basic idea is: since it is generally believed that the dynamic system of piezoelectric micro-actuator is high-order continuous differentiable, the approximate linear system of the form of formula (1) exists. Then for such a system, two adjacent measurement states within a period of time must be approximately related by the following formula:
[0078] x k+1 ≈Ax k +Bu k (2)
[0079] If the data is measured at a constant sampling interval Δt, then x k+1 , denote the state variables at time (k+1)Δt and time kΔt respectively, Represents the control input at time kΔt.
[0080] Assume that for a piezoelectric microactuator with unknown dynamics, a total of s+1 state variables and their corresponding s control inputs are collected in a short period of time. These sampled data are stored in the queue defined in S2.1 and recorded as:
[0081] X=[x1,x2,…,x s ],X'=[x2,x3,…,x s+1 ],Υ=[u1,u2,...,u s]
[0082] Then from formula (2) we can get:
[0083] X'≈AX+BY (3)
[0084] in Furthermore, let G = [AB], Ω = [X T Y T ] T , then formula (3) can be rewritten as:
[0085] X'=GΩ (4)
[0086] Then G can be obtained by the following formula:
[0087]
[0088] in is a pseudo-inverse operator. The matrices A and B as sub-matrices of the block matrix G can be indirectly obtained through equation (5). Substituting the obtained A and B into equation (2), equation (2) is called the discrete linear approximation dynamic system obtained by the piezoelectric micro-actuator at the current state.
[0089] The dynamic linearization algorithm of the piezoelectric microactuator is as follows:
[0090]
[0091]
[0092] Step 3: Design a model-free predictive controller based on online data according to the linear approximation model, and substitute the reference trajectory, system matrix A, control input matrix B, current state x(k), and weight matrix Q, W into the model-free predictive controller to obtain the optimal control sequence;
[0093] Based on the above embodiment, step 3 specifically includes:
[0094] Step 3.1, design a model-free predictive controller based on online data according to the linear approximation model and transform it into a constrained quadratic programming problem;
[0095] Step 3.2, calculate the parameters of the quadratic programming problem based on the reference trajectory, system matrix A, control input matrix B, current state x(k) and weight matrix Q, W;
[0096] Step 3.3, after substituting all parameters, solve the quadratic programming problem by the interior point method to obtain the optimal control sequence.
[0097] Furthermore, the expression of the quadratic programming problem is
[0098]
[0099]
[0100] Among them, the parameters H,f of the quadratic programming problem T The definition is as follows:
[0101] H=2(θQθ+W),f T =2E T Qθ
[0102] E=Ψx(k)-R k , Where A and B are the system matrix and control input matrix of the linear approximation dynamics system, p is the control time domain length of the model predictive control, x(k) is the system state at the current time k, R k is the reference trajectory sequence.
[0103] In specific implementation, the key to the design of the model-free predictive controller based on online data in the present invention is: in step 2, a dynamic linearization method based on dynamic modal decomposition with control terms is selected, and after the linear approximation model of the piezoelectric micro-actuator is solved through the system input and output sampling data, a model predictive controller is designed according to the linear approximation model.
[0104] Let x(k+i|k), u(k+i|k), and r(k+i|k) represent the predicted values of the system state and control input at time k+i and the expected value of the system state obtained based on the linear approximation model and the system state x(k) at the current time k. And define the following symbols:
[0105] X k =[x(k+1|k),x(k+2|k),…,x(k+p|k)] T
[0106] U k =[u(k+1|k),u(k+2|k),…,u(k+p|k)] T
[0107] R k =[r(k+1|k),r(k+2|k),…,r(k+p|k)] T
[0108] Where p is the control time domain length of the model predictive control, X k is the system state prediction sequence with a length of p starting from the kth moment, U k is the control input sequence of length p starting from the kth moment, R k is a reference trajectory sequence of length p starting from the kth moment.
[0109] Suppose the feasible set of control input is Then the model predictive control based on the linear approximation model can be formalized as the following constrained quadratic programming problem:
[0110]
[0111] The definitions of the parameters in this quadratic programming problem are as follows:
[0112] H=2(θQθ+W),f T =2E T Qθ,E=Ψx(k)-R k
[0113]
[0114] The system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the approximation system obtained in step S2 can be used to calculate the parameters of the quadratic programming problem. After substituting the parameters, the above quadratic programming problem (6) can be solved by the interior point method, and an optimal control sequence U can be obtained. k .
[0115] In summary, the model-free predictive controller algorithm based on online data is as follows:
[0116]
[0117] Step 4, using the first control variable of the optimal control sequence as the optimal control input for the next step to control the piezoelectric micro-actuator control system.
[0118] In specific implementation, after obtaining the optimal control sequence, U k The first control quantity U k (1) As the optimal control input u for the next step * , and then the optimal control input is used to control the piezoelectric microactuator control system.
[0119] The piezoelectric micro-actuator model-free predictive control method based on online data provided in this embodiment collects the I / O data of the system and constructs an approximation solver based on the dynamic modal decomposition method with control items to solve the linear approximation dynamic system, thereby realizing the discrete dynamic linearization of the unknown dynamic system. Then, a model predictive controller is designed based on the linear approximation model to perform ultra-precision motion control on the piezoelectric micro-actuator. While achieving similar model-free predictive control performance, the controller itself has lower computing performance requirements and is more suitable for use in embedded controllers with weak computing power.
[0120] The present solution will be described below in conjunction with a specific embodiment. In this embodiment, the control object is a one-dimensional piezoelectric micro-actuator. In the simulation experiment of the embodiment, a dynamic model is needed to simulate an actual one-dimensional piezoelectric micro-actuator. The dynamic model is as follows:
[0121]
[0122] where x, represents the displacement, velocity and acceleration of the piezoelectric microactuator, Its control input, represents the hysteresis nonlinear term in its dynamics. m,b,k, are model parameters, m is the equivalent mass of the piezoelectric microactuator, b is its equivalent damping coefficient, k is its equivalent elastic coefficient, d e In this embodiment, the dynamics of the hysteresis term h will be described using the following Bouc-Wen model:
[0123]
[0124] Among them, α, β, are the parameters of the Bouc-Wen model. The dynamic model of the piezoelectric microactuator is composed of equations (7) and (8).
[0125] In this implementation example, the model-free predictive control method based on online data does not contain any information about the piezoelectric micro-actuator dynamics model. This dynamics model is only used to simulate the real piezoelectric micro-actuator to obtain sampling data. The model parameters set in the simulation experiment of the implementation example are derived from [7]:
[0126] m=0.28kg, b=1302.28N / (m / s), k=53452N / m, de=0.127nm / V,
[0127] α=0.5136, β=0.124, λ=-0.073
[0128] S1. Given the desired reference trajectory r(t)
[0129] The one-dimensional piezoelectric microactuator used in this embodiment has two state variables and one control input, that is, n = 2, m = 1. represents the true value of the system state variable at time t, Represents the true value of the system control input at time t.
[0130] The setting control is to track a fixed trajectory, which is a special case of trajectory tracking control. In this embodiment, the desired trajectory of the setting control is r s(t) = 1 μm. The desired trajectory of trajectory tracking control is r t (t) = 1 + 0.1sin(100t) μm. Let ε(t) represent the displacement tracking error of the system.
[0131] S2. Dynamic Linearization of Piezoelectric Microactuators
[0132] Solving the linear approximation dynamics of the piezoelectric microactuator in real time is called dynamic linearization. In this implementation example, we will use the dynamic modal decomposition method with control terms to solve the linear approximation dynamics. The steps are as follows:
[0133] S2.1 Define the data structure of the sampled data
[0134] In order to realize the dynamic linearization of the piezoelectric microactuator, in the actual algorithm, it is necessary to first define three queues, X, X', and Y. These three queues follow the first-in-first-out rule when updating the data elements therein: when a new data is sampled, the new sampled data can only enter from the head of the queue, and at the same time, an old data is taken out from the tail of the queue and deleted.
[0135] S2.2 Input the initial control signal I(t) for initial sampling
[0136] Before the control system starts running, since the input and output data of the system have not yet been obtained, it is necessary to give a small initial input signal and collect s groups of initial input and output sampling data. Assuming that the sampling interval of the system is Δt, the initial input signal needs to last at least T = (s-1) Δt. The initial signal set in this implementation example is Then the external input of the closed-loop control system is recorded as:
[0137]
[0138] Let x(t) and u(t) represent the system state variables and external control input values of the piezoelectric micro-actuator at different times. After the initial sampling is completed, an initial queue storing the initial sampling data will be obtained as shown below:
[0139] X0=[x(0),x(Δt),…,x((s-1)Δt)],
[0140] X'0=[x(Δt),x(2Δt),…,x(sΔt)],
[0141] Y0=[u(0),u(Δt),…,u((s-1)Δt)]
[0142] These initial sampling data will be used to derive the initial linear approximation dynamics of the unknown system.
[0143] S2.3 Solving the linear approximation dynamics
[0144] In this embodiment, the sampling interval is set to Δt=1e-6s, the number of sampling nodes is set to s=10, and the input of Algorithm 1 is n=2, m=1, s=10, Δt=1e-6s. Algorithm 1 can be used to solve the linear approximation dynamic parameters A and B of the piezoelectric microactuator.
[0145] S3 Design of Model-Free Predictive Controller Based on Online Data
[0146] In this embodiment, a dynamic linearization method based on dynamic modal decomposition with control terms is selected in step S2 to solve the linear approximation model. The key to the design of a model-free predictive controller based on online data is: after solving the linear approximation model of the piezoelectric micro-actuator through the system input and output sampling data, a model predictive controller is designed according to the linear approximation model.
[0147] Let x(k+i|k), u(k+i|k), r(k+i|k) represent the predicted value of the system state and control input at time k+i and the expected value of the system state obtained based on the linear approximation model and the system state x(k) at the current time k. In this implementation example, the control time domain length of the model predictive control is set to p=5. And the following symbols are defined:
[0148] X k =[x(k+1|k),x(k+2|k),...,x(k+5|k)] T
[0149] U k =[u(k+1|k),u(k+2|k),...,u(k+5|k)] T
[0150] Rs k =[r s (k+1|k),r s (k+2|k),...,r s (k+5|k)] T
[0151] R k =[r t (k+1|k),r t (k+2|k),...,r t (k+5|k)] T
[0152] Where X k is the system prediction state sequence with a length of 5 starting from the kth moment. kis a control input sequence of length 5 starting from the kth moment. k is the reference trajectory sequence of the control setting starting from the kth moment with a length of 5, Rt k is a trajectory tracking control reference trajectory sequence with a length of 5 starting from the kth moment.
[0153] In this implementation example, it is assumed that the minimum value of the control input voltage is 0 V and the maximum value is 50 V. In addition, the control input u in the actual application scenario cannot change suddenly, so it is assumed that the control input also needs to meet the following Lipschitz continuity condition:
[0154] ‖u(t+Δt)-u(t)||≤L u ‖Δt‖
[0155] Where L u is the Lipschitz continuity constant. In this embodiment, let L u =100000. Then the feasible set of control input is Then the model predictive control based on the linear approximation model can be formalized as the following constrained quadratic programming problem:
[0156]
[0157]
[0158] in:
[0159] H=2(θQθ+W),f T =2E T Qθ,E=Ψx(k)-R k
[0160]
[0161] The weight matrices Q and W in the optimization problem of this implementation example are set as follows:
[0162]
[0163]
[0164] The system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the approximation system obtained in step S2 can be used to calculate the parameters of the quadratic programming problem. After substituting the parameters, the above quadratic programming problem (6) can be solved by the interior point method, and an optimal control sequence U can be obtained. k , and finally k The first control quantity U k (1) As the optimal control input u for the next step *In this implementation example, the quadprog function in MATLAB can be directly called to solve the above quadratic programming problem using the interior point method. k =Rs k When * is the optimal control input for setting control. k =Rt k When * is the optimal control input for trajectory tracking control.
[0165] S4 applies the model-free predictive control method based on online data to piezoelectric microactuators and conducts simulation experiments
[0166] The model-free predictive control method based on online data is applied to the piezoelectric micro-actuator. The mathematical software MATLAB is used to perform simulation experiments on the setting control and trajectory tracking control, and the experimental data are collected and plotted. The simulation time is set to 0.1 seconds. Figure 4 As shown in FIG. 1 , (a) is the setting control effect of the present invention when applied to an implementation example, and (b) is its control error. Figure 5 As shown, (a) is the trajectory tracking control effect when the present invention is applied to an implementation example, and (b) is its control error. From this, it can be seen from a macroscopic perspective that under the model-free predictive control based on online data, the displacement of the piezoelectric microactuator quickly converges to a given reference trajectory. From a microscopic perspective, it can be seen that the steady-state accuracy of the set tracking reaches |ε|≤1.1e-2μm, and the steady-state mean square error of the trajectory tracking control is 1.13e-2μm. This shows that when the present invention is applied to a piezoelectric microactuator, both the set control and the trajectory tracking control have very high control accuracy.
[0167] It should be understood that various parts of the present disclosure may be implemented in hardware, software, firmware, or a combination thereof.
[0168] The above is only a specific implementation of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present disclosure should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.
Claims
1. A model-free predictive control method for piezoelectric microactuators based on online data, It is characterized in that include: Step 1, giving the desired reference trajectory of the piezoelectric microactuator control system; Step 2, using the dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric micro-actuator, and obtain the linear approximation model corresponding to the piezoelectric micro-actuator, as well as the system matrix A, control input matrix B, current state x(k) and weight matrix Q, W of the linear approximation dynamic system; The step 2 specifically includes: Step 2.1, define the data structure of the sampling data; Step 2.2, inputting an initial control signal I(t) to perform initial sampling on the piezoelectric microactuator; Step 2.3, after the initial sampling is completed, the sampled data is used to form a linear approximation dynamic system of the decompressed electric microactuator; Step 2.4, using a dynamic mode decomposition method with control terms to solve the linear approximation dynamic system of the piezoelectric microactuator, and obtaining a linear approximation model corresponding to the piezoelectric microactuator; The step 2.1 specifically includes: Define multiple queues corresponding to the sampled data and the data update rules for each queue; The step 2.2 specifically includes: Define the sampling interval and calculate the sampling period based on the preset sampling quantity; Starting from inputting the initial control signal I(t), the system state variables of the piezoelectric micro-actuator and the values of the external control input at different times are sampled within the sampling period; Step 3: Design a model-free predictive controller based on online data according to the linear approximation model, and substitute the reference trajectory, system matrix A, control input matrix B, current state x(k), and weight matrix Q, W into the model-free predictive controller to obtain the optimal control sequence; The step 3 specifically includes: Step 3.1, design a model-free predictive controller based on online data according to the linear approximation model and transform it into a constrained quadratic programming problem; Step 3.2, calculate the parameters of the quadratic programming problem based on the reference trajectory, system matrix A, control input matrix B, current state x(k) and weight matrix Q, W; Step 3.3, after substituting all parameters, solve the quadratic programming problem by the interior point method to obtain the optimal control sequence; The expression of the quadratic programming problem is Among them, the parameters H,f of the quadratic programming problem T The definition is as follows: H=2(θQθ+W),f T =2E T Qθ Where A and B are the system matrix and control input matrix of the linear approximation dynamics system, p is the control time domain length of the model predictive control, x(k) is the system state at the current time k, R k is the reference trajectory sequence, U k is the control input sequence of length p starting from the kth moment, is the feasible set of control inputs; Step 4, using the first control variable of the optimal control sequence as the optimal control input of the next step to control the piezoelectric micro-actuator control system.
