System identification input design method and device for wireless power transmission system

By adopting the input sequence design method of rolling time domain in the radio energy transmission system, the problem that traditional sequences cannot fully stimulate the system is solved, the identification accuracy and calculation efficiency are improved, and faster input optimization is achieved.

CN119089691AInactive Publication Date: 2024-12-06WUHAN UNIV

Patent Information

Application Number
CN202411211442.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-12-06
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the existing radio energy transmission systems, traditional pseudo-random binary sequences (PRBS sequences) cannot fully stimulate the system, resulting in low system identification accuracy, high complexity of the optimization algorithm for input optimization problems and long calculation time.

Method used

Using the input sequence design method based on the rolling time domain, the D optimization criteria and Taylor expansion are used to reduce the number of variables to be optimized, and the optimal input is quickly calculated using the SQP algorithm to improve the D index of the information matrix, so that the input can more fully stimulate the system mode. state, enrich the dynamic information output by the system.

Benefits of technology

It improves the model identification accuracy of the radio energy transmission system, reduces the complexity of the objective function, reduces the calculation time, and achieves faster input optimization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089691A_ABST
    Figure CN119089691A_ABST
Patent Text Reader

Abstract

The invention provides a system identification input design method and device for a wireless power transmission system, and the method comprises the steps: obtaining input and output data, determining the G order of a model, enabling the G order of a transfer function model to be optimized according to a D criterion, taking the maximum determinant of an M information matrix as a target, building a target function J, simplifying the target function J, and carrying out the design of the system identification input of the wireless power transmission system. Calculating a target function J (k) at the moment k and writing the target function J (k) into an input optimization module of the upper computer, starting the input optimization module and an identification module of the upper computer, recursively identifying a system model at each sampling moment by the identification module, calculating the optimal input at the moment k + 1 by the input optimization module based on the target function J (k), and the calculation of optimal input is realized by adopting a sequential quadratic programming method. According to the method, a rolling time domain D optimization method is adopted, an objective function of input optimization is determined and simplified, and the optimization complexity is greatly reduced; the SQP algorithm based on the Hessian matrix is adopted to solve the input optimization problem, and the calculation speed can be greatly increased.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless power transmission, and in particular to a system identification input design method and device for a wireless power transmission system. Background Art

[0002] Wireless power transmission systems have been widely used in household and industrial fields such as smartphones, wearable devices and electric vehicles. In order to design a controller to achieve control requirements such as constant current / constant voltage, it is necessary to model the wireless power transmission system to analyze the dynamic and static control characteristics of the system. At present, a data-driven system identification method can be used to establish a control model for the wireless power transmission system. That is, a suitable excitation sequence is applied to the input control quantity (such as the inverter phase shift angle) and the response sequence of the system output quantity (such as the load voltage at the receiving end) is measured, and the system mathematical model is fitted based on the data.

[0003] In order to reflect the real state of the wireless charging system and improve the accuracy of identification, the experimental data needs to contain as much behavioral information of the system characteristics as possible. Therefore, it is necessary to design the identification input signal. The current input optimization method has the following disadvantages:

[0004] I. The problem cannot be directly converted into a multi-stage decision problem, so the objective function for input optimization cannot be determined;

[0005] II. The optimization algorithm for solving the input optimization problem has high complexity and long computation time. Summary of the invention

[0006] In view of the technical problems existing in the prior art, the present invention provides a system identification input design method and device for a wireless power transmission system, which is used to solve the problem that the traditional pseudo-random binary sequence (PRBS sequence) in the identification of the wireless power transmission system cannot fully excite the system. An input sequence design method based on the rolling time domain idea is proposed, which can make the information matrix reach D optimal, make the identified model parameters closer to the true value, and improve the model identification accuracy.

[0007] According to a first aspect of the present invention, the present invention provides a system identification input design method for a wireless power transmission system, comprising the following steps:

[0008] Step 1. Obtain input and output data, observe the output waveform, and determine the G order of the transfer function model;

[0009] Step 2. The transfer function model G order is optimized according to the D optimization criterion, and the objective function J is established with the maximum determinant of the M information matrix as the goal;

[0010] Step 3. Based on the rolling time domain principle, Taylor expansion is performed on the M information matrix at time k, and the objective function J is converted into the information matrix increment at time k, thereby reducing the number of variables to be optimized and simplifying the objective function J;

[0011] Step 4. Collect the input, output and identification results at time k, pass them into the prediction model, calculate the objective function J(k) at time k, and write it into the input optimization module of the host computer;

[0012] Step 5. Connect the wireless power transmission system to the host computer, start the input optimization module and identification module of the host computer, the identification module recursively identifies the system model at each sampling time, and the input optimization module calculates the optimal input at time k+1 based on the objective function J(k) at time k.

[0013] Based on the above technical solution, the present invention can also make the following improvements.

[0014] Optionally, the acquiring input and output data, observing the output waveform, and determining the G order of the transfer function model includes:

[0015] After applying a step signal with a small amplitude at the equilibrium point of the system input end, the system output curve is collected to observe the characteristics of the system output curve. If the output curve does not oscillate, the first-order transfer function model is used to fit the current system; when the output curve oscillates, the second-order or higher-order transfer function model is used to fit the current system.

[0016] Optionally, the G-order transfer function model is optimized according to the D optimization criterion, and the objective function J is established with the maximum determinant of the M information matrix as the goal, including:

[0017] Convert the continuous transfer function model G after the order is determined into a discrete ARX model, and write out the parameter estimates according to the least squares algorithm According to the D optimization criterion, the objective function J is established with the goal of maximizing the determinant of the M information matrix.

[0018] Optionally, based on the rolling time domain principle, Taylor expansion is performed on the M information matrix at time k, and the objective function J is converted to obtain the information matrix increment at time k, thereby reducing the number of variables to be optimized. The simplified objective function J includes:

[0019] According to the prediction time domain nh, the input to be optimized {u(k+1),u(k+2),…,u(k+n h )}, Taylor expansion is performed on the information matrix M(k) to obtain the information matrix increment at time k, and the simplified objective function J at time k is obtained as:

[0020]

[0021] Optionally, the input, output and identification results at time k are collected, passed into the prediction model, and the objective function J(k) at time k is calculated, including:

[0022] Assuming that the current time is k, collect the input u(k), output y(k) and identification result θ(k) at this time, pass them into the prediction model, and calculate Then, the objective function J(k) at time k is calculated according to formula (4) and formula (5).

[0023] Optionally, the identification module recursively identifies the system model at each sampling moment, and the identification equation is as follows:

[0024]

[0025] in represents the identified parameter vector, λ represents the forgetting factor, represents the regression vector, and P(k) represents the identification process vector.

[0026] Optionally, in step five, the wireless power transmission system and the host computer are connected through a data transceiver device, and the data transceiver device transmits the collected system output data to the host computer. The host computer transmits the designed input to the data transceiver device chip through the communication interface, and then sends it to the FPGA module to generate a switch tube signal for adjusting the input phase shift angle.

[0027] Optionally, before starting the input optimization module and identification module of the host computer, it is also necessary to set the initial parameters θ0 of the identification model and the input upper and lower limits [u min ,u max ]parameter.

[0028] According to a second aspect of the present invention, there is provided a wireless power transmission system identification input device, comprising: a wireless power transmission system, a data transceiver and a host computer, wherein:

[0029] The host computer includes an input optimization module and an identification module, which are used to generate input signals and identify system models, and design input signals and identify system models according to collected experimental data;

[0030] The data transceiver connects the wireless power transmission system and the host computer, and sends the data collected by the wireless power transmission system to the host computer for processing, or transmits the data generated by the host computer to the wireless power transmission system.

[0031] Optionally, the data transceiver device collects the output of the wireless power transmission system and uses the output phase shift angle control signal as the system input, and the host computer optimizes the wireless power transmission system input based on the collected data.

[0032] Technical effects and advantages of the present invention:

[0033] The present invention provides a system identification input design method and device for a wireless power transmission system. The present invention adopts a rolling time domain D optimization input design method to establish an optimization problem based on sampling data and identification results at the current moment, thereby reducing the complexity of the objective function and the number of inputs to be optimized. The SQP algorithm is used to quickly calculate the optimal input at the next moment to improve the D index of the model information matrix, so that the input can more fully stimulate the system mode, enrich the dynamic information contained in the system output, and ultimately improve the identification accuracy of the wireless power transmission system.

[0034] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a structural diagram of a D optimization algorithm provided by an embodiment of the present invention;

[0036] Figure 2 is a rolling time domain principle diagram provided by an embodiment of the present invention;

[0037] Figure 3 is a system input and output diagram provided by an embodiment of the present invention;

[0038] Figure 4 It is a structural diagram of a device for wireless power transmission system identification and input optimization provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] It should be noted that the embodiment of the present invention aims at the system identification of the mathematical model of the wireless power transmission system, and proposes to calculate the optimal input at the next moment by collecting data at the current moment. The embodiment of the present invention mainly solves the problem that the traditional pseudo-random binary sequence (PRBS sequence) in the identification of the wireless power transmission system cannot fully excite the system, and proposes an input sequence design method based on the rolling time domain idea, which can make the information matrix reach D optimal, make the model parameters obtained by identification closer to the true value, and improve the model identification accuracy.

[0041] Another purpose of an embodiment of the present invention is to solve how to establish and simplify the objective function for calculating the optimal input: a D optimization objective function design method based on a rolling time domain is provided, the objective function form is determined according to the D optimization index, and the objective function in the rolling time domain form is further calculated, thereby greatly reducing the dimension of the optimization problem.

[0042] Another purpose of the embodiment of the present invention is to solve the problem of calculating the objective function: a sequential quadratic programming method (SQP method) based on the Hessian matrix is ​​provided to solve the objective function. The algorithm has a fast calculation speed and can directly solve the discrete objective function, which can greatly improve the input optimization efficiency.

[0043] It can be understood that based on the defects in the background technology, the embodiment of the present invention proposes a system identification input design method for a wireless power transmission system, comprising the following steps:

[0044] Step 1. Obtain input and output data, observe the output waveform, and determine the G order of the transfer function model;

[0045] In this embodiment, the obtaining of input and output data, observing the output waveform, and determining the G order of the transfer function model specifically includes:

[0046] After applying a step signal with a small amplitude at the equilibrium point of the system input end, the system output curve is collected to observe the characteristics of the system output curve. If the output curve does not oscillate, the first-order transfer function model is used to fit the current system; when the output curve oscillates, the second-order or higher-order transfer function model is used to fit the current system.

[0047] The specific goal of identification is the small signal transfer function G(s,θ) of the system at the operating point, where s represents the differential operator and θ represents the parameter vector to be identified.

[0048] Therefore, the specific application scenarios of the G-order model are determined as follows: first, a step signal with a small amplitude is applied to the balance point of the system input (such as the inverter phase shift angle) to obtain the system output (such as the load voltage) curve. The characteristics of the output curve are observed on the host computer. When the output curve does not oscillate, the first-order model is used to fit the current system; when the output curve oscillates, the second-order or higher-order model is used to fit the current system.

[0049] Step 2. The transfer function model G order is optimized according to the D optimization criterion, and the objective function J is established with the maximum determinant of the M information matrix as the goal;

[0050] In this embodiment, the transfer function model G order, according to the rolling time domain D optimization criterion, the objective function J is established with the maximum determinant of the M information matrix as the goal, including:

[0051] Convert the continuous transfer function model G after the order is determined into a discrete ARX model, and write out the parameter estimation values ​​according to the principle of the least squares algorithm According to the D optimization criterion, the objective function J is established with the goal of maximizing the determinant of the M information matrix.

[0052] It should be noted that the D optimization algorithm structure is as follows Figure 1 As shown in the figure, it includes the WPT to be identified, the identification module, the prediction model and the D optimization module. The least squares identification module recursively identifies the system model at each sampling moment. The prediction model calculates the predicted output y(k+1)…y(k+nh) based on the predicted input u(k+1)…u(k+nh) according to the latest identification parameters, and passes it to the input optimization module to calculate the optimal input u*(k+1) at the next moment.

[0053] In the steps of this embodiment, after determining the order of the transfer function model G, the continuous transfer function model G is converted into a discrete ARX model suitable for least squares identification (RLS). After the ARX model is written in a form linear with respect to the parameters ε(k) represents noise, assuming its mean is 0 and covariance is a constant σ 2 The regression vector required for identification can be determined The objective function of least squares identification is So the estimated value of parameter θ is The expression is:

[0054]

[0055] Where φ(k) and Y(k) are from 1 to k and y(i), i=1,2,…,k:

[0056]

[0057] Y(k)=[y(1),y(2),...,y(k)] T

[0058] The covariance matrix of the least squares identification is used to evaluate the accuracy of the identification and is defined as follows:

[0059]

[0060] where θ * Represents the true value of the parameter. In order to improve the identification accuracy, that is, to minimize the covariance matrix, this problem can be equivalent to maximizing the information matrix

[0061] According to the D optimization criterion, the objective function is established with the maximum determinant of the M information matrix as the goal. At the same time, in order to eliminate the influence of the sampling data length on the matrix size, the input optimization problem can be written as:

[0062]

[0063] Where N represents the length of the sampled data, and the determinant value of the information matrix is ​​maximized by optimizing the input u from time 1 to N.

[0064] Step 3. Based on the rolling time domain principle, Taylor expansion is performed on the M information matrix at time k, and the objective function J is converted into the information matrix increment at time k, thereby reducing the number of variables to be optimized and simplifying the objective function J;

[0065] Since the number of inputs to be optimized is consistent with the data length, directly solving equation (3) to optimize the input requires a huge amount of computation, so it is necessary to reduce the number of inputs to be optimized.

[0066] The embodiment of the present invention intends to optimize the increment of the information matrix at each sampling moment based on the rolling time domain concept, thereby reducing the number of inputs to be optimized and simplifying the optimization problem.

[0067] The schematic diagram of rolling time domain is as follows Figure 2 As shown in the figure, when the system runs to time k, the system input u(k) and system output y(k) are collected to calculate the identification parameter vector θ(k). According to the determined prediction time domain length nh and the optimized objective function J, the optimized input vector u(k+1),…,u(k+nh) is calculated. Then u(k+1) is injected into the system for system identification at the next moment.

[0068] According to the prediction time domain nh, the input to be optimized {u(k+1),u(k+2),…,u(k+n h)}, and then Taylor expand the information matrix M(k) to obtain the information matrix increment at time k, so the simplified objective function J at time k is written as:

[0069]

[0070] Write the calculated objective function J into the host computer.

[0071] Step 4. Collect the input, output and identification results at time k, pass them into the prediction model, calculate the objective function J(k) at time k, and write it into the input optimization module of the host computer;

[0072] Calculate the objective function J(k) at time k: Assume that the current time is k, collect the input u(k), output y(k) and identification result θ(k) at this time, pass them into the prediction model, and calculate Then, the objective function J(k) at time k is calculated according to (4) and (5);

[0073] Step 5. Connect the wireless power transmission system to the host computer, start the input optimization module and identification module of the host computer, the identification module recursively identifies the system model at each sampling time, and the input optimization module calculates the optimal input at time k+1 based on the objective function J(k) at time k.

[0074] In this embodiment, the wireless power transmission system is connected to the host computer through a data transceiver device, and the data transceiver device transmits the collected system output data to the host computer. The host computer transmits the designed input to the data transceiver device chip through the communication interface, and then sends it to the FPGA module to generate a switch tube signal to adjust the input phase shift angle. When setting the initial parameters of the identification model θ0, the upper and lower limits [u min ,u max ] parameters, start the input optimization module and identification module of the host computer. The input optimization module will calculate the optimal input at the next moment at each sampling moment, and the identification module will recursively identify the system model at each sampling moment. The identification equation is as follows:

[0075]

[0076] in, represents the identified parameter vector, λ represents the forgetting factor, It represents the regression vector, and its specific form is determined by the structure of the model G and the input and output collected by the system; P(k) represents the identification process vector.

[0077] The embodiment of the present invention proposes a system identification input design method for a wireless power transmission system, determines the objective function form according to the D optimization index, and based on the rolling time domain idea, calculates the objective function based on the sampled data and the identification result at each moment, and takes the system input of the next few steps as the variable to be optimized, thereby greatly reducing the dimension of the optimization problem; adopts the sequential quadratic programming method (SQP method) based on the Hessian matrix to solve the objective function, and the algorithm can directly solve the discrete objective function, thereby quickly obtaining the optimal input.

[0078] Next, a typical string-to-string compensation wireless power transmission system is used as an example for specific description. The method includes the following steps:

[0079] Step 1: Determine the model G order. Identify the small signal transfer function from the inverter phase shift angle (system input u) to the load voltage (system output y). First determine the current operating point of the system And the operating frequency fs. When the static operating point of the system Set to When fs=100kHz, the static output voltage of the system is 378V. A step sequence is applied to the system input (i.e., the inverter phase shift angle), that is, Set to The data is transmitted to the STM32 chip through the RS485 module. The STM32 module transmits the signal to the FPGA chip on the circuit board to adjust the phase shift angle u(k) of the inverter. The collected voltage y(k) is transmitted to the STM32 chip using ST-LINK. Then the data is converted and transmitted to the host computer. The voltage waveform is generated using Matlab software. The waveform shows that when the system input step changes, the system output does not show obvious oscillation. Therefore, the first-order system G can be used to fit the system:

[0080]

[0081] Where s represents the differential operator, a and b represent the parameters to be identified, θ = [a, b] T represents the identified parameter vector, and x represents the system output without noise.

[0082] Step 2: Identify the regression variables According to step 1, the identification model is a first-order model, and the number of parameters to be identified is 2, so the regression vector to be identified can be determined The order of is 2, the prediction time domain nh is set to nh = 3, and the input to be optimized at time k is {u(k+1), u(k+2), u(k+3)}. In order to use the least squares algorithm to identify the transfer functions a and b, it is necessary to convert (4) into a discrete form. The present invention uses a multi-step integration method to rewrite (4) as follows:

[0083]

[0084] in,

[0085]

[0086] in, In this experiment, Ts is 1 ms.

[0087] Step 3: Determine the form of the objective function J. According to the optimization criterion D, that is, write the objective function J with the maximum determinant of the information matrix M as the target column, and write the objective function according to (4) and (5):

[0088]

[0089] in And write it to the host computer.

[0090] Step 4: Calculate the objective function J(k) at time k; at time k, the input u(k) and output y(k) collected by the STM32 chip are transmitted to the host computer, and the host computer passes the input u(k) and output y(k) and the identification result θ(k) into the prediction model, and substitutes {u(k+1),u(k+2),u(k+3)} into the prediction model to calculate Then, the objective function J(k) at time k is calculated according to (10).

[0091] Step 5, wireless power transmission system identification based on input optimization: connect the wireless power transmission system to the host computer through the data transceiver. Set the initial parameter of the identification model θ(0) = 1e -4 ×I2 (I2 represents the second-order unit matrix), identification process parameters P(0) = 1e4 × I2, forgetting factor λ = 0.99, sampling time Ts = 1ms. Start the input optimization module and identification module of the host computer, and at the working point of the system A small perturbation sequence {u(1), ..., u(L)} is applied to excite the system mode, where L (L = 5000) represents the total length of the input sequence. The input optimization module calculates the optimal input {u(k+1), u(k+2), u(k+3)} at each sampling time k, and then injects its first term u(k+1) into the system. The identification module identifies the system model at each sampling time according to the recursive equation (7):

[0092]

[0093] According to the optimized input and output data Finally, it can be identified that the wireless power transmission system Small signal transfer function model when

[0094] Figure 3 The output of the identification model is compared with the actual output. The upper part of the picture shows the waveform of the system input phase shift angle, and the lower part shows the comparison between the actual output obtained by sampling (blue solid line) and the simulation output (red dotted line) obtained after adding the system input based on the identified model. It can be seen that the output of the identified model can fit the actual sampled output well, thus proving the accuracy of the identification model.

[0095] In summary, the embodiment of the present invention optimizes the system input based on the rolling horizon D optimization concept, improves the system model identification accuracy, and verifies the correctness of the system through experiments.

[0096] In addition, an embodiment of the present invention provides a wireless power transmission system identification input device, comprising: a wireless power transmission system, a data transceiver and a host computer, wherein:

[0097] The host computer includes an input optimization module and an identification module, which are used to generate input signals and identify system models, and design input signals and identify system models according to collected experimental data;

[0098] The data transceiver connects the wireless power transmission system and the host computer, and sends the data collected by the wireless power transmission system to the host computer for processing, or transmits the data generated by the host computer to the wireless power transmission system.

[0099] The data transceiver collects the output of the wireless power transmission system and uses the output phase shift angle control signal as the system input, and the host computer optimizes the wireless power transmission system input based on the collected data.

[0100] The identification input device can send the data collected on the system to the host computer for processing through the data transceiver device, and can also transmit the data generated by the host computer to the wireless power transmission system. Figure 4The figure shows the schematic diagram of the structure of the wireless power transmission system identification input device, which mainly includes a data transceiver and a host computer. The data transceiver STM32 transmits the system output data collected by ST-LINK to the host computer, and the host computer transmits the designed input to the data transceiver STM32 chip through the communication interface RS485, and then sends it to the FPGA module to generate the switch tube signal; the host computer includes an input optimization module and an identification module, which can design the input signal and identify the system model according to the collected experimental data. The function of the data transceiver includes collecting the system output and the output phase shift angle control signal as the system input. The host computer optimizes the system input based on the collected data, thereby improving the system identification accuracy. The identification input device also includes: a DC power supply, an inverter circuit, a primary side series compensation circuit, a secondary side series compensation circuit, a rectifier filter circuit, a load, a data acquisition device and an identification and input optimization device; in the figure, the main parameters of the wireless power transmission system are shown in Table 1 below.

[0101] Table 1 Wireless power transmission system parameters

[0102]

[0103]

[0104] In summary, a wireless power transmission system identification input device provided by an embodiment of the present invention optimizes the system input based on the rolling time domain D optimization concept, improves the system model identification accuracy, and verifies the correctness of the system through experiments.

[0105] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A system identification input design method for a wireless power transmission system, characterized in that: The following steps are involved: Step 1. Obtain input and output data, observe the output waveform, and determine the G order of the transfer function model; Step 2. The transfer function model G order is optimized according to the D optimization criterion, and the objective function J is established with the maximum determinant of the M information matrix as the goal; Step 3. Based on the rolling time domain principle, Taylor expansion is performed on the M information matrix at time k, and the objective function J is converted into the information matrix increment at time k, thereby reducing the number of variables to be optimized and simplifying the objective function J; Step 4. Collect the input, output and identification results at time k, pass them into the prediction model, calculate the objective function J(k) at time k, and write it into the input optimization module of the host computer; Step 5. Connect the wireless power transmission system to the host computer, start the input optimization module and identification module of the host computer, the identification module recursively identifies the system model at each sampling time, and the input optimization module calculates the optimal input at time k+1 based on the objective function J(k) at time k.

2. The system identification input design method of a wireless power transmission system according to claim 1, characterized in that: The obtaining of input and output data, observing the output waveform, and determining the G order of the transfer function model comprises: After applying a step signal with a small amplitude at the equilibrium point of the system input end, the system output curve is collected to observe the characteristics of the system output curve. If the output curve does not oscillate, the first-order transfer function model is used to fit the current system; when the output curve oscillates, the second-order or higher-order transfer function model is used to fit the current system.

3. The system identification input design method of a wireless power transmission system according to claim 1, characterized in that: The transfer function model G order is optimized according to the D optimization criterion, and the objective function J is established with the maximum determinant of the M information matrix as the goal, including: Convert the continuous transfer function model G after the order is determined into a discrete ARX model, and write out the parameter estimates according to the least squares algorithm According to the optimization criterion D, the objective function J is established with the goal of maximizing the determinant of the information matrix M.

4. The system identification input design method of a wireless power transmission system according to claim 1, characterized in that: Based on the rolling time domain principle, the M information matrix is ​​Taylor expanded at time k, the objective function J is converted into the information matrix increment at time k, the number of variables to be optimized is reduced, and the simplified objective function J includes: According to the prediction time domain nh, the input to be optimized {u(k+1),u(k+2),…,u(k+n h )}, Taylor expansion is performed on the information matrix M(k) to obtain the information matrix increment at time k, and the simplified objective function J at time k is obtained as:

5. The system identification input design method of a wireless power transmission system according to claim 4, characterized in that: The input, output and identification results at time k are collected and passed into the prediction model to calculate the objective function J(k) at time k, including: Assuming that the current time is k, collect the input u(k), output y(k) and identification result θ(k) at this time, pass them into the prediction model, and calculate Then, the objective function J(k) at time k is calculated according to formula (4) and formula (5).

6. A system identification input design method for a wireless power transmission system according to claim 1, characterized in that: The identification module recursively identifies the system model at each sampling moment, and the identification equation is as follows: in, represents the identified parameter vector, λ represents the forgetting factor, represents the regression vector, and P(k) represents the identification process vector.

7. The system identification input design method of a wireless power transmission system according to claim 1, characterized in that: In step five, the wireless power transmission system and the host computer are connected through a data transceiver device, and the data transceiver device transmits the collected system output data to the host computer. The host computer transmits the designed input to the data transceiver chip through the communication interface, and then sends it to the FPGA module to generate a switch tube signal for adjusting the input phase shift angle.

8. The system identification input design method of a wireless power transmission system according to claim 1, characterized in that: Before starting the input optimization module and identification module of the host computer, it is also necessary to set the initial parameters θ0 of the identification model and the input upper and lower limits [u min ,u max ]parameter.

9. A wireless power transmission system identification input device, used in a wireless power transmission system system identification input design method according to any one of claims 1 to 8, characterized in that: The device comprises: a wireless power transmission system, a data transceiver and a host computer, wherein: The host computer includes an input optimization module and an identification module, which are used to generate input signals and identify system models, and design input signals and identify system models according to collected experimental data; The data transceiver connects the wireless power transmission system and the host computer, and sends the data collected by the wireless power transmission system to the host computer for processing, or transmits the data generated by the host computer to the wireless power transmission system.

10. The wireless power transmission system identification input device according to claim 9, characterized in that: The data transceiver is used to collect the output of the wireless power transmission system and use the output phase shift angle control signal as the system input, and the host computer optimizes the wireless power transmission system input based on the collected data.

Citation Information

Patent Citations

  • Piezoelectric nanometer positioning system control method and device based on rolling optimization, electronic equipment and storage medium

    CN115933385A

  • Method and device for controlling iron horse in channel

    CN118170032A

Cited By

  • SQP-based wind power gear box order-only model parameter identification method

    CN121088579A