Parameter identification technology for magneto-rheological damper mechanical model

Abstract

The invention discloses a parameter identification method for a magneto-rheological damper mechanical model. The parameter identification method comprises the following steps: step 1, testing and obtaining an output curve of a magneto-rheological damper; step 2, a Bouc-wen model is built in MATLAB (matrix laboratory) / SIMULINK (single matrix laboratory); step 3, performing parameter optimization processing by using a particle swarm optimization algorithm as an optimization tool, including: determining a target function as # imgabs0 #; step 4, judging the recognition precision of the optimization algorithm; 5, performing polynomial fitting; 6, evaluating the output condition of the model; according to the method, the problem of complex nonlinear output and mathematical model parameter identification of the magneto-rheological damper can be converted into a process parameter identification process, the process is simple, the requirement for priori knowledge of engineers is not high, and the method has important guiding significance for practical engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of vibration control, and more particularly to a parameter identification technology for a mechanical model of a magnetorheological damper, which can be applied, for example, to the semi-active vibration control of magnetorheological dampers in the fields of earthquake, wind vibration, civil and military vehicles, and artillery launch engineering. Background Art

[0002] In recent years, with the increasing requirements and control objectives for vibration control in the fields of construction engineering, equipment, and military industry, traditional passive vibration isolation technologies are difficult to adapt to the changes of vibration sources. Although active control technologies have good effects, their designed systems are complex and energy consumption is large. Therefore, engineers and scholars have begun to seek a class of methods between passive and active control - semi-active control, and semi-active control based on magnetorheological dampers is the most common method. The parameter identification of magnetorheological dampers is the key to accurately and effectively carrying out semi-active control. It has important practical significance to adopt what kind of method to carry out the parameter identification of magnetorheological dampers. Summary of the Invention

[0003] In view of the above problems, the present invention proposes a parameter identification technology for a mechanical model of a magnetorheological damper. Based on the mechanical experiments of the magnetorheological damper, the output force conditions of the magnetorheological damper under different frequencies, different amplitudes, and different current conditions are obtained; then the swarm intelligence optimization algorithm - particle swarm optimization algorithm is introduced, which has the advantages of simple structure, convenient parameter setting, high calculation efficiency, etc.; by proposing an effective objective fitness function, the parameter optimization process is made more rapid and effective; in addition, the parameter optimization range is also optimized one by one, and finally a relatively accurate and effective parameter optimization range is obtained, ensuring the high efficiency and accuracy of parameter optimization; based on the obtained identified parameters, polynomial fitting of different currents and the identified parameters is carried out, aiming to predict the model parameters under different currents, and accurate identification of 8 unknown parameters of the mechanical model of the magnetorheological damper is achieved.

[0004] More specifically, according to one aspect of the present invention, there is provided a method for parameter identification of a mechanical model of a magnetorheological damper, including:

[0005] Step 1, test and obtain the output force curve of the magnetorheological damper under a harmonic vibration load, different vibration frequencies, different vibration amplitudes, and different currents or voltages;

[0006] Step 2, build a Bouc-wen model in MATLAB / SIMULINK, and the mechanical expressions are shown as formulas (1) and (2):

[0007]

[0008] In the formula, F represents the controllable damping force of the magnetorheological damper; x represents the relative displacement of the magnetorheological damper; represents the relative velocity of the magnetorheological damper; represents the differential of the virtual state variable z; A, β, γ, α, c0, k0, n, and x0 are 8 unknown parameters to be identified;

[0009] Step 3: Using the particle swarm optimization algorithm as an optimization tool, perform parameter optimization, including:

[0010] Set the initial values of the model parameters;

[0011] Determine the objective function as where m is the number of test data points, F Fit and F Exp are the damping forces obtained by simulation and test through the particle swarm optimization algorithm, respectively; is the difference between the maximum and minimum damping force values in the test data;

[0012] Set the initial search ranges for the 8 unknown parameters A, β, γ, α, c0, k0, n, and x0, including the lower and upper limits of the search ranges;

[0013] According to the optimization results, adjust the upper limits of the search for the 8 unknown parameters respectively until the results cannot be further optimized, and determine the final parameter search ranges;

[0014] According to the determined parameter search ranges, use the particle swarm optimization algorithm to optimize and identify the 8 unknown parameters; and

[0015] After reaching the accuracy rate of the target fitness value, export the optimized and identified parameters;

[0016] Step 4: According to the identified parameters, substitute them back into the Bouc-wen model, and evaluate the error between the output force obtained from the identified model and the actual output force to judge the accuracy of the optimization algorithm identification;

[0017] Step 5: Perform polynomial fitting on the 8 identified parameters under different currents;

[0018] Step 6: According to the mathematical expression obtained from the polynomial fitting, calculate and predict the Bouc-wen model parameters under other currents, and perform output force calculation to evaluate the output force situation of the model.

[0019] According to the implementation of the present invention, wherein said Step 1 further includes determining the accuracy of the data obtained from the vibration table test of the magnetorheological damper, and the output force data is directly related to the accuracy of parameter identification.

[0020] According to the implementation of the present invention, in Step 1, the output force curve includes the force-velocity curve and the force-displacement curve.

[0021] According to an embodiment of the present invention, in step three, the parameter setting is initialized through the statistics of the observed data or empirical estimation.

[0022] According to an embodiment of the present invention, the order of the polynomial fitting should not exceed the 6th order.

[0023] According to another aspect of the present invention, there is provided a parameter identification device for a mechanical model of a magnetorheological damper, including:

[0024] Output Curve Test Module of Magneto-Rheological Damper , for testing and obtaining the output force curve of the magnetorheological damper under the action of a harmonic vibration load, different vibration frequencies, different vibration amplitudes, different currents or voltages;

[0025] Bouc-wen Model Building Module For building a Bouc-wen model in MATLAB / SIMULINK, the mechanical expression is represented by equations (1) and (2):

[0026]

[0027] In the formula, F represents the controllable damping force of the magnetorheological damper; x represents the relative displacement of the magnetorheological damper; represents the relative velocity of the magnetorheological damper; represents the differential of the virtual state variable z; A, β, γ, α, c0, k0, n, and x0 are 8 unknown parameters to be identified;

[0028] Parameter Optimization Processing Module For using the particle swarm optimization algorithm as an optimization tool to perform parameter optimization processing, including:

[0029] Set the initial values of the model parameters;

[0030] Determine the objective function as where m is the number of test data points, F Fit and F Exp are the damping forces obtained by simulation and experiment through the particle swarm optimization algorithm respectively; is the difference between the maximum and minimum damping force values in the test data;

[0031] Set the initial search ranges of 8 unknown parameters A, β, γ, α, c0, k0, n, and x0, including the lower and upper limits of the search ranges;

[0032] According to the optimization results, adjust the upper limits of the search of the 8 unknown parameters respectively until the results cannot be further optimized, and determine the final parameter search ranges;

[0033] Search for and identify 8 unknown parameters using the particle swarm optimization algorithm according to the determined parameter search range; and

[0034] After reaching the accuracy rate of the target fitness value, export the parameters identified by the optimization search;

[0035] An accuracy judgment module for the optimization algorithm identification, which is used to substitute the identified parameters back into the Bouc-Wen model and evaluate the error between the output obtained from the identified model and the actual output to judge the accuracy of the optimization algorithm identification;

[0036] A polynomial fitting module, which is used to perform polynomial fitting on the 8 identified parameters under different currents;

[0037] A model output situation evaluation module, which is used to calculate and predict the Bouc-Wen model parameters under other currents according to the mathematical expression obtained by polynomial fitting, and perform output calculation to evaluate the output situation of the model.

[0038] According to another aspect of the present invention, there is also provided an electronic device, including: a memory and one or more processors;

[0039] The memory is used to store one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the present invention.

[0040] The present invention makes the parameter optimization process more rapid and effective by proposing an effective target fitness value function (objective function); in addition, the parameter optimization range is also optimized design by design, and finally a relatively accurate and effective parameter optimization range is obtained, ensuring the efficiency and accuracy of parameter optimization; on the basis of obtaining the identified parameters, polynomial fitting of different currents and the identified parameters is performed, aiming to predict the model parameters under different currents, and accurate identification of 8 unknown parameters of the magnetorheological damper mechanical model is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the semi-active control technology process of a magnetorheological damper for strong shock vibration control according to an embodiment of the present invention;

[0042] Figure 2 It is a schematic diagram of the Bouc-Wen model structure of the semi-active control technology of a magnetorheological damper for strong shock vibration control according to an embodiment of the present invention;

[0043] Figure 3 It is a calculation flow chart of the Bouc-Wen model of the semi-active control technology of a magnetorheological damper for strong shock vibration control according to an embodiment of the present invention;

[0044] Figure 4 Comparison diagram of predicted and actual output forces of magnetorheological dampers under different currents for the semi-active control technology of magnetorheological dampers for strong shock and vibration control according to an embodiment of the present invention

[0045] Figure 5 Schematic structural diagram of a parameter identification device for a magnetorheological damper mechanical model according to an embodiment of the present invention; and

[0046] Figure 6 Schematic structural diagram of an electronic device according to an embodiment of the present invention Detailed implementation manners

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and through specific embodiments. The content shown is used to fully elaborate the content of the present invention and does not limit the present invention

[0048] It should be understood that models and tools such as the Bouc-Wen model, MATLAB / SIMULINK, particle swarm optimization algorithm, polynomial fitting, etc. involved in the present invention are known per se. Therefore, the present invention focuses on how to combine and optimize the above various tools or models to design the process of the parameter identification technology for the magnetorheological damper mechanical model of the present invention

[0049] Figure 1 Schematic flow diagram of the semi-active control technology of magnetorheological dampers for strong shock and vibration control according to an embodiment of the present invention. Refer to Figure 1 , the parameter identification method for the magnetorheological damper mechanical model according to an embodiment of the present invention may include:

[0050] First, use the magnetorheological damper for testing to obtain the mechanical data of the magnetorheological damper. For example, obtain the output force curves of the magnetorheological damper under harmonic vibration loads, different vibration frequencies, different vibration amplitudes, and different currents (or voltages), including the force-velocity curve and the force-displacement curve. The output data is directly related to the accuracy of parameter identification. Therefore, the accuracy of the data obtained by the magnetorheological damper after the vibration table test should be ensured

[0051] Step two, build a Bouc-Wen model in MATLAB / SIMULINK Figure 2 Schematic structural diagram of the Bouc-Wen model of the semi-active control technology of magnetorheological dampers for strong shock and vibration control according to an embodiment of the present invention Figure 3 Calculation flow chart of the Bouc-Wen model of the semi-active control technology of magnetorheological dampers for strong shock and vibration control according to an embodiment of the present invention

[0052] For the Bouc-Wen model, the mechanical expression can be represented as: and The determined parameters to be identified are eight unknown parameters: A, β, γ, α, c0, k0, n, and x0. Among them, F represents the controllable damping force of the magnetorheological damper; x represents the relative displacement of the magnetorheological damper; represents the relative velocity of the magnetorheological damper; represents the differential of the virtual state variable z; c0 is the viscous damping coefficient; k0 is the stiffness coefficient; x0 is the initial displacement deviation; α is the ratio of the stiffness before and after yielding; z is the hysteresis variable; γ, β, and A are model parameters determined by the magnetorheological fluid and the control system; n is a coefficient.

[0053] Step 3: After building the Bouc - wen model, use the particle swarm optimization algorithm as an optimization tool to optimize the parameters of the model.

[0054] Specifically, the optimization process may include:

[0055] Set initial values for the model parameters, which can be initialized through statistical analysis of the observed data or empirical estimation.

[0056] Determine the objective function as where m is the number of experimental data points, F Fit and F Exp are the damping forces obtained through simulation by the particle swarm optimization algorithm and through experiments, respectively; is the difference between the maximum and minimum damping force values in the experimental data. The fitness function (objective function) proposed in the present invention is comprehensively selected by the inventor based on an accurate understanding of the output mechanism of the magnetorheological damper and combined with engineering and optimization calculation experience. This fitness function has a good optimization effect on the parameters to be identified.

[0057] Set the initial search ranges for the eight unknown parameters A, β, γ, α, c0, k0, n, and x0, including the lower and upper limits of the search ranges; the definition of the initial search ranges can generally be determined according to prior knowledge. Adjust the upper limits of the search for the eight unknown parameters respectively according to the optimization results until the results cannot be further optimized, and determine the final parameter search ranges; according to the determined parameter search ranges, use the particle swarm optimization algorithm to optimize and identify the eight unknown parameters; after reaching the accuracy rate of the target fitness value, export the optimized and identified parameters, thus completing the optimization and identification process: Table 1 below shows the identification results of the Bouc - Wen model parameters under different currents according to the embodiments of the present invention. This table will guide the polynomial fitting of the identified parameters with different currents and provide a basis for predicting the model parameters under other different currents.

[0058] Table 1: Identification Results of Bouc - Wen Model Parameters under Different Currents

[0059]

[0060] Step 4: According to the identified parameters, substitute them back into the Bouc-Wen model, and evaluate the error between the output obtained from the identification model and the actual output to determine the accuracy of the optimization algorithm identification. Figure 4 It is a comparison diagram of the predicted output and the actual output of the magnetorheological damper for the semi-active control technology of the magnetorheological damper for strong shock vibration control according to the implementation scheme of the present invention under different currents.

[0061] Step 5: After obtaining the identified parameters under different currents, perform polynomial fitting on the 8 identified parameters under different currents to predict the parameters under other currents. This is a step for the present invention to have good applicability, realizing the accurate and comprehensive identification of the 8 unknown parameters of the mechanical model of the magnetorheological damper. In addition, the order of polynomial fitting is generally not more than 6.

[0062] Finally, according to the mathematical expression obtained by polynomial fitting, calculate and predict the Bouc-Wen model parameters under other currents, and perform output calculation to evaluate the output of the model.

[0063] Figure 5 It is a schematic diagram of the parameter identification device for the mechanical model of the magnetorheological damper according to the implementation scheme of the present invention. As Figure 5 shown, the device includes: an output curve test module 210 of the magnetorheological damper, which is used to test and obtain the output curve of the magnetorheological damper under the action of a harmonic vibration load, different vibration frequencies, different vibration amplitudes, different currents or voltages; a Bouc-Wen model building module 220, which is used to build a Bouc-Wen model in MATLAB / SIMULINK; a parameter optimization processing module 230, which is used to use the particle swarm optimization algorithm as an optimization tool to perform parameter optimization processing; an accuracy judgment module 240 for optimizing algorithm identification: which is used to substitute the identified parameters back into the Bouc-Wen model and evaluate the error between the output obtained from the identification model and the actual output to determine the accuracy of the optimization algorithm identification; a polynomial fitting module 250, which is used to perform polynomial fitting on the 8 identified parameters under different currents; a model output situation evaluation module 260, which is used to calculate and predict the Bouc-Wen model parameters under other currents according to the mathematical expression obtained by polynomial fitting, and perform output calculation to evaluate the output situation of the model.

[0064] Figure 6 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. As Figure 6 shown, the electronic device includes a processor 310, a memory 320, an input device 330, and an output device 340; the number of processors 310 in the electronic device can be one or more.Figure 3 Take a processor 310 in [the device] as an example; the processor 310, memory 320, input device 330, and output device 340 in the electronic device can be connected through a bus or other means. Figure 6 Take the connection through a bus as an example.

[0065] The memory 320, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the parameter identification method for the mechanical model of the magnetorheological damper in the embodiments of the present invention (for example, the output force curve test module 210 of the magnetorheological damper; the Bouc-wen model building module 220; the parameter optimization processing module 230; the accuracy judgment module 240 for identifying the optimization algorithm; the polynomial fitting module 250; the model output situation evaluation module 260). The processor 310 executes various functional applications and data processing of the electronic device by running the software programs, instructions, and modules stored in the memory 320, that is, implements the above-mentioned parameter identification method for the mechanical model of the magnetorheological damper.

[0066] The parameter identification technology for the mechanical model of the magnetorheological damper of the present invention is ingeniously designed and has prominent key points. It can start from the mechanism of the output force of the magnetorheological damper and propose a parameter identification method widely applicable to various types of magnetorheological dampers. This method is simple to implement, closely combines swarm intelligence optimization with parameter identification, and ensures the effect of parameter identification by setting a clever fitness function and determining an effective parameter search range. It has strong applicability and operability. The invention can transform the problem of complex non-linear output force and complex mathematical model parameter identification of the magnetorheological damper into a process-based parameter identification process, which is simple and has low requirements for the prior knowledge of engineers, and has important guiding significance for practical engineering.

[0067] The above description of the embodiments is for the ordinary technical personnel in the technical field to understand and apply the present invention. Those familiar with the technology in the field can obviously easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative labor. Therefore, the present invention is not limited to the embodiments here, and the improvements and modifications made by those skilled in the art without departing from the scope of the present invention should be within the protection scope of the present invention.

Claims

1. A parameter identification method for a magnetorheological damper mechanical model, characterized in that: include: Step 1, testing and obtaining the output curve of the magnetorheological damper under simple harmonic vibration load, different vibration frequencies, different vibration amplitudes, and different currents or voltages; Step 2: Build the Bouc-wen model in MATLAB / SIMULINK, and the mechanical expressions are expressed as equations (1) and (2): Wherein, F represents the controllable damping force of the magnetorheological damper; x represents the relative displacement of the magnetorheological damper; represents the relative velocity of the magnetorheological damper; represents the differential of the virtual state variable z; A, β, γ, α, c0, k0, n and x0 are a total of 8 unknown parameters to be identified; Step 3: Use the particle swarm optimization algorithm as an optimization tool to perform parameter optimization, including: Set the initial values ​​of model parameters; Determine the objective function as Where m is the number of test data points, F Fit and F Exp are the damping forces obtained by simulation and experiment through particle swarm optimization algorithm, respectively; is the difference between the maximum and minimum damping force values ​​in the test data; Set the initial search range of 8 unknown parameters, namely A, β, γ, α, c0, k0, n and x0, including the lower and upper limits of the search range; Adjust the search upper limits of the eight unknown parameters according to the optimization results until the results cannot be further optimized, and determine the final parameter search range; Based on the determined parameter search range, the particle swarm optimization algorithm is used to identify the eight unknown parameters; and After reaching the target fitness value accuracy, the optimal identification parameters are derived; Step 4: According to the identified parameters, re-substitute into the Bouc-wen model and evaluate the error between the output obtained according to the identified model and the actual output to determine the accuracy of the optimization algorithm identification; Step 5: Perform polynomial fitting on the 8 identification parameters under different currents; Step 6: Based on the mathematical expression obtained by polynomial fitting, calculate and predict the Bouc-wen model parameters under other currents, and perform output calculation to evaluate the output of the model.

2. The parameter identification method for a magnetorheological damper mechanical model according to claim 1, characterized in that: The step 1 also includes determining the accuracy of the data obtained after the magnetorheological damper is tested on a vibration table, and the output data is directly related to the accuracy of parameter identification.

3. The parameter identification method for the mechanical model of a magnetorheological damper according to claim 1, characterized in that: In step 1, the output curve includes a force-velocity curve and a force-displacement curve.

4. The parameter identification method for a magnetorheological damper mechanical model according to claim 1, characterized in that: In step three, parameter settings are initialized through statistical or empirical estimation of observed data.

5. The parameter identification method for a magnetorheological damper mechanical model according to claim 1, characterized in that: The polynomial fitting order should not exceed 6.

6. A parameter identification device for a magnetorheological damper mechanical model, characterized in that: include: The output curve test module of the magnetorheological damper is used to test and obtain the output curve of the magnetorheological damper under the action of simple harmonic vibration load, different vibration frequencies, different vibration amplitudes, different currents or voltages; The Bouc-wen model building module is used to build the Bouc-wen model in MATLAB / SIMULINK. The mechanical expressions are expressed as equations (1) and (2): Wherein, F represents the controllable damping force of the magnetorheological damper; x represents the relative displacement of the magnetorheological damper; represents the relative velocity of the magnetorheological damper; represents the differential of the virtual state variable z; A, β, γ, α, c0, k0, n and x0 are a total of 8 unknown parameters to be identified; The parameter optimization processing module is used to use the particle swarm optimization algorithm as an optimization tool to perform parameter optimization processing, including: Set the initial values ​​of model parameters; Determine the objective function as Where m is the number of test data points, F Fit and F Exp are the damping forces obtained by simulation and experiment through particle swarm optimization algorithm, respectively; is the difference between the maximum and minimum damping force values ​​in the test data; Set the initial search range of 8 unknown parameters, namely A, β, γ, α, c0, k0, n and x0, including the lower and upper limits of the search range; Adjust the search upper limits of the eight unknown parameters according to the optimization results until the results cannot be further optimized, and determine the final parameter search range; Based on the determined parameter search range, the particle swarm optimization algorithm is used to identify the eight unknown parameters; and After reaching the target fitness value accuracy, the optimal identification parameters are derived; The precision judgment module of the optimization algorithm identification is used to re-substitute the Bouc-wen model according to the identified parameters and evaluate the error between the output obtained according to the identified model and the actual output to judge the precision of the optimization algorithm identification; and The polynomial fitting module is used to perform polynomial fitting on the eight identification parameters under different currents. The model output evaluation module is used to calculate and predict the Bouc-wen model parameters under other currents based on the mathematical expressions obtained by polynomial fitting, and perform output calculation to evaluate the output of the model.

7. An electronic device, characterized in that: include: memory and one or more processors; The memory is used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

Images