Evaluation method of mechanical model of magnetorheological damper based on fuzzy comprehensive evaluation method

Abstract

This scheme provides an evaluation method for the mechanical model of magnetorheological vibration damper based on the fuzzy comprehensive evaluation method, including: S1: determining indicators; S2: determining the objective weights of secondary indicators; S3: calculating the comprehensive score of the primary indicators based on the objective weights of the secondary indicators and the original sample data values; S4: determining the subjective weights of the primary indicators based on the hierarchical analysis method, and obtaining a quantitative score of the mechanical model of magnetorheological vibration damper based on the weighted comprehensive score of the primary indicators based on the subjective weights. The coupling system of the fuzzy comprehensive evaluation method and the hierarchical analysis method is applied to the evaluation of the mechanical model of magnetorheological vibration damper to address the shortcomings of the existing evaluation methods in terms of systematicity and objectivity.

Description

Technical Field

[0001] The present invention relates to the field of vehicle intelligent suspension model evaluation, and in particular to a magnetorheological damper mechanical model evaluation method based on a fuzzy comprehensive evaluation method. Background Art

[0002] Magnetorheological suspension is a special type of semi-active suspension. Its core mechanism is to precisely control the damping coefficient of the shock absorber by regulating the rheological properties of the magnetorheological fluid through the intensity of the magnetic field, thereby optimizing vehicle dynamics. The key actuator of the magnetorheological suspension is the magnetorheological damper. Its operating mechanism is based on the phase change characteristics of the magnetorheological fluid under an external magnetic field: in the zero-field state, it exhibits Newtonian fluid properties, but under magnetic field excitation, it transforms into a Bingham fluid. The magnetic particles form a directional chain structure, which significantly enhances the shear yield stress, thereby achieving real-time controllable adjustment of the damping force.

[0003] Due to the inherent nonlinear and hysteretic properties of magnetorheological dampers, the dynamic modeling process of magnetorheological dampers is significantly complex. To accurately predict and control their dynamic responsiveness, a high-precision model that can accurately characterize the nonlinear characteristics of the damper and analyze the hysteretic dissipation mechanism is required. Existing modeling methods can be divided into two major paradigms: parametric and non-parametric. Parametric models (such as the Bingham model and the Bouc-Wen model) characterize the dynamic behavior of magnetorheological dampers based on physical parameters. Their construction requires a comprehensive consideration of parameter accuracy, experimental data reliability, and model structure complexity. Non-parametric models (such as neural networks and fuzzy logic models) fit the mechanical response through a data-driven approach. When dealing with time-varying nonlinear systems, they can effectively circumvent the mathematical representation limitations of traditional physical constitutive models and exhibit greater adaptability to working conditions and modeling flexibility.

[0004] The evaluation of the mechanical model of the magnetorheological shock absorber is a core link in the research and development and application of magnetorheological suspension systems. Its accuracy directly determines the effectiveness of the control strategy, serves as a bridge between theory and experiment, and is the key to driving the improvement of suspension system performance. Specifically, a high-precision mechanical model of the magnetorheological shock absorber can accurately predict the dynamic response of the damping force, providing a reliable foundation for various control algorithms and avoiding the deterioration of suspension performance due to model errors. The evaluation process ensures the physical meaning of the parametric model and the generalization ability of the non-parametric model through experimental verification, ensuring the accuracy of the mechanical model of the magnetorheological shock absorber under different working conditions. At the same time, the mechanical model of the magnetorheological shock absorber can quantify multi-objective conflicts, helping the magnetorheological suspension to achieve a balance between comfort, handling stability and durability, and supporting multi-scale modeling collaboration, reducing the cost of physical prototype iteration, and providing full-chain support for the research and development and application of magnetorheological suspension systems.

[0005] Validation of the mechanical model of a magnetorheological damper requires a systematic evaluation. Existing evaluation systems primarily rely on metrics such as relative error, root mean square error (RMS), hysteresis loop shape, and computational time. However, these systems often suffer from a single dimensionality and a lack of dynamic characterization. For example, Patent 201810376477.8 employs a dynamic hysteresis unit model, validating the model through the relative error of the damping force. Patent 202411365685.X uses a Gaussian process regression model and measures accuracy through RMS error. While these single-parameter statistical methods can quantify static deviations, they struggle to analyze dynamic response characteristics. Another example is the composite polynomial model proposed in Patent 200810155989.8, which qualitatively evaluates the dynamic characteristics based on hysteresis curve shape. While this model can intuitively reflect dynamic characteristics, its lack of quantitative criteria can easily introduce subjective errors. Other studies use computational efficiency as an evaluation criterion, but this emphasizes computational speed while neglecting prediction accuracy.

[0006] In summary, the current evaluation system for the mechanical model of magnetorheological dampers has not yet established a multi-dimensional collaborative analysis framework. It is urgent to establish a multi-criteria comprehensive evaluation method that integrates model accuracy, dynamic characteristics resolution, and computational efficiency. The construction of this system needs to solve complex problems such as multi-objective optimization and indicator weight allocation. Summary of the Invention

[0007] The purpose of the present invention is to provide a magnetorheological damper mechanical model evaluation method based on the fuzzy comprehensive evaluation method, and to apply the fuzzy comprehensive evaluation method and the hierarchical analysis method coupling system to the evaluation of the magnetorheological damper mechanical model to solve the shortcomings of the existing evaluation methods in terms of systematicity and objectivity.

[0008] To achieve the above objectives, the present technical solution provides a method for evaluating the mechanical model of a magnetorheological damper based on a simulation comprehensive evaluation method, comprising the following steps:

[0009] S1: Determine the indicators:

[0010] Determine the primary index of the magnetorheological shock absorber and the secondary index corresponding to the primary index;

[0011] S2: Determine the objective weight of the secondary indicators;

[0012] Obtain the original sample data values of the secondary indicators of the magnetorheological damper under different frequencies and currents, and process the original sample data values using the entropy method and the coefficient of variation method to obtain the secondary indicator weight matrix of the secondary indicators under each primary indicator;

[0013] S3: Calculate the comprehensive score of the first-level indicators based on the objective weights of the second-level indicators and the original sample data values:

[0014] Based on the global interval of the original sample data value of each secondary indicator, different evaluation levels are obtained, the evaluation level of each secondary indicator is converted into a secondary indicator score vector, the secondary indicator fuzzy relationship matrix is obtained based on the evaluation level and the original sample data value of the secondary indicator, the secondary indicator fuzzy relationship matrix is weighted according to the secondary indicator weight matrix to obtain the secondary indicator fuzzy comprehensive evaluation matrix, and the scalar product operation is performed on the secondary indicator fuzzy relationship matrix and the secondary indicator score vector to obtain the first-level indicator comprehensive score;

[0015] S4: Determine the subjective weight of the first-level indicators based on the hierarchical analysis method, and obtain the quantitative score of the magnetorheological shock absorber mechanical model based on the comprehensive score of the first-level indicators weighted by the subjective weight.

[0016] Compared with the existing technology, this technical solution has the following characteristics and beneficial effects:

[0017] This application scheme constructs a mechanical model evaluation system for magnetorheological shock absorbers: first, a three-dimensional evaluation framework including statistical parameters, hysteresis characteristics and operating performance is established, and model fidelity, model business value and model software driving quality are specifically set as first-level indicators, and seven second-level indicators such as damping force root mean square error, indicator diagram morphological parameters, and operating time are used to obtain an evaluation model. The second-level indicators are processed by adopting the entropy weight method and the coefficient of variation combined weighting method, and the objective weights of the second-level indicators are obtained through the weight fusion mechanism of information entropy difference quantification and data discreteness representation, so as to solve the defects of inconsistent dimensions and insufficient high correlation of the second-level indicators related to the mechanical model due to multi-factor coupling, nonlinear relationship and parameter uncertainty in the modeling process of magnetorheological shock absorbers. At the same time, the analytic hierarchy process is combined with demand-oriented subjective empowerment of the first-level indicators to construct a weight distribution system that is coordinated between subjective and objective factors. Finally, the quantitative score of the computational mechanics model is integrated with the constructed evaluation model through membership function conversion, realizing the normalized representation of multi-dimensional performance indicators and the visual comparison of cross-model differences. This coupled evaluation system provides quantitative decision-making support for the modeling and optimization of complex nonlinear systems of magnetorheological dampers. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of the magnetorheological damper mechanical model evaluation method based on the fuzzy comprehensive evaluation method provided in this scheme.

[0019] Figure 2 It is an indicator system of primary and secondary indicators of the magnetorheological shock absorber mechanical model evaluation method based on fuzzy comprehensive evaluation method provided by this scheme.

[0020] Figure 3 This is a logical diagram of calculating the comprehensive score of the first-level indicators of the magnetorheological shock absorber mechanical model evaluation method based on the fuzzy comprehensive evaluation method provided in this scheme. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention are within the scope of protection of the present invention.

[0022] It should be understood by those skilled in the art that, in the disclosure of the present invention, the terms "longitudinal", "transverse", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the above terms should not be understood as limiting the present invention.

[0023] Example 1

[0024] like Figure 1 and Figure 3 As shown, this scheme provides a magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method, which includes the following steps:

[0025] S1: Determine the indicators:

[0026] The model fidelity, business value, and software-driven quality of the magnetorheological damper are used as primary indicators. The root mean square error and coefficient of determination of the model damping force are used as secondary indicators of model fidelity. The relative rate of change of the total area of the model dynamometer diagram, the relative rate of change of the area ratio of the dynamometer diagram, and the relative rate of change of the saturation of the dynamometer diagram are used as secondary indicators of model business value. The model running time and model memory usage are used as secondary indicators of model software-driven quality.

[0027] S2: Determine the objective weight of the secondary indicators;

[0028] Obtain the original sample data values of the secondary indicators of the magnetorheological damper under different frequencies and currents, and process the original sample data values using the entropy method and the coefficient of variation method to obtain the secondary indicator weight matrix of the secondary indicators under each primary indicator;

[0029] S3: Calculate the comprehensive score of the first-level indicators based on the objective weights of the second-level indicators and the original sample data values:

[0030] Based on the global range of the original sample data values of each secondary indicator, different evaluation levels are obtained, and the evaluation level of each secondary indicator is converted into a secondary indicator score vector; according to the ratio of the number of original sample data values of each secondary indicator that meets the evaluation level to the total number of original sample data values of the secondary indicator, the membership of the secondary indicator under the evaluation level is used as the ratio, the membership of each secondary indicator under different evaluation levels is sorted to obtain a membership matrix, and the membership matrices of the secondary indicators under the same primary indicator are spliced to obtain a secondary indicator fuzzy relationship matrix; according to the secondary indicator weight matrix, the secondary indicator fuzzy relationship matrix is weighted to obtain a secondary indicator fuzzy comprehensive evaluation matrix, and the secondary indicator fuzzy relationship matrix and the secondary indicator score vector are scalar-producted to obtain a comprehensive score of the primary indicator;

[0031] S4: Determine the subjective weight of the first-level indicators based on the hierarchical analysis method, and obtain the quantitative score of the magnetorheological shock absorber mechanical model based on the comprehensive score of the first-level indicators weighted by the subjective weight.

[0032] Regarding step S1 of this protocol:

[0033] like Figure 2 As shown, this approach selects model fidelity, model business value, and model software driver quality as first-level indicators when evaluating the mechanical model of a magnetorheological damper. Corresponding second-level indicators are selected for each first-level indicator. Specifically, the second-level indicators corresponding to model fidelity are the root mean square error (RMSE) and the coefficient of determination (R²) between the predicted and experimental values of the mechanical model's damping force; the second-level indicators corresponding to model business value are the relative rate of change of the model's total area, the relative rate of change of the area ratio of the model's dynamometer diagram, and the relative rate of change of the saturation of the model's dynamometer diagram; and the second-level indicators corresponding to model software driver quality are the model's runtime and the model's memory usage.

[0034] It should be noted that the systematic construction of indicators for the mechanical model of magnetorheological shock absorbers is the core foundation for ensuring the scientific nature of the evaluation system and the credibility of the results. Its selection must take into account comprehensiveness, objectivity and characterization capabilities in order to accurately characterize the intrinsic characteristics and external manifestations of the evaluation object. To comprehensively reflect the accuracy, stability, and computational efficiency of the mechanical model of the magnetorheological damper, this proposal selects the model fidelity, model business value, and model software driver quality of the magnetorheological damper as first-level indicators, taking into account the fitting accuracy, hysteresis curve characteristics, and model operation resource consumption of statistical parameters. Model fidelity is used to characterize the mechanical model's fitting accuracy and ability to restore the actual mechanical characteristics of the magnetorheological damper, reflecting the reliability of the mechanical model at the statistical level. Model business value is used to characterize the practical value and performance of the mechanical model in engineering applications, focusing on the mechanical model's ability to represent the physical meaning of the dynamic characteristics of the damper. Model software driver quality characterizes the operating efficiency and resource utilization characteristics of the mechanical model in the software environment, reflecting the computational feasibility and adaptability of the mechanical model in practical applications. Therefore, the three first-level indicators selected in this proposal can characterize the theoretical fit, engineering applicability, and system implementation efficiency of the magnetorheological damper mechanical characteristic model from different dimensions.

[0035] The first-level indicators of this plan are:

[0036] U={U1,U2,U3};

[0037] Among them, U1, U2, and U3 are first-level indicators regarding model fidelity, business application value, and software-driven quality, respectively.

[0038] Furthermore, in view of the incompleteness of a single statistical parameter, this scheme determines that the secondary indicators corresponding to model fidelity are the root mean square error (RMSE) and coefficient of determination (R²) of the model damping force. The root mean square error (RMSE) of the model damping force measures the degree of deviation between the predicted value of the mechanical model (such as the damping force) and the actual measured value. The smaller the RMSE, the lower the degree of dispersion between the predicted value of the mechanical model and the actual data, and the higher the fitting accuracy. The coefficient of determination (R²) of the simulated damping force reflects the mechanical model's ability to explain data variation. The closer R² is to 1, the more comprehensive the mechanical model's description of the mechanical properties of the shock absorber and the stronger the fidelity. The set of secondary indicators for model fidelity is expressed as:

[0039] U1={u 11 ,u 12};

[0040] where u 11 ,u 12 are the root mean square error (RMSE) and coefficient of determination (R²) of the model damping force, respectively.

[0041] ;

[0042] ;

[0043] Where RMSE is the root mean square error of the model damping force, R² is the coefficient of determination of the model damping force, t is the number of damping force samples of the model; F dl is the experimental value of the damping force of the magnetorheological damper; is the fitting value of the damping force; is the mean damping force.

[0044] In order to avoid the subjectivity of the hysteresis curve shape judgment of the mechanical model, the secondary indicators corresponding to the business value of the model are determined to be the relative change rate of the total area of the model dynamometer diagram, the relative change rate of the area ratio of the dynamometer diagram, and the relative change rate of the saturation of the dynamometer diagram. The set of secondary indicators of the business value of the model is expressed as:

[0045] U2={u 21 ,u 22， u 23};

[0046] where u 21 ,u 22， u 23 They correspond to the relative change rate of the total area, area ratio and saturation of the indicator diagram respectively.

[0047] The dynamometer diagram is a closed hysteresis curve used to characterize the relationship between the damping force and displacement of the magnetorheological shock absorber. The dynamometer diagram is divided into two parts, the upper and lower parts, with the damping force being zero as the dividing line. The areas of these two parts are recorded as S 上 and S 下 In addition, the maximum difference in damping force is taken as the length, and the maximum difference in displacement is taken as the width. The circumscribed rectangle of the dynamometer curve is constructed, and its area is recorded as S 矩 , based on this, the total area of the dynamometer diagram is S 上 and S 下 The sum of the area ratio of the indicator diagram is S 上 and S 下 The saturation of the dynamometer diagram is the ratio of the total area of the dynamometer diagram to S 矩 The relative change rate of the dynamometer diagram saturation refers to the relative change rate of the dynamometer diagram saturation fitted by the mechanical model compared to the dynamometer diagram saturation of the test value of the bench test, that is, the change rate of the fitted value compared to the actual test value.

[0048] Taking into account the resource consumption and efficiency of the mechanical model operation, the secondary indicators corresponding to the model software driver quality are model running time and model memory usage. The secondary indicator set of the model software driver quality is expressed as:

[0049] U3={u 31 ,u 32};

[0050] where u 31 ,u 32 They correspond to the model running time and the model memory occupied respectively.

[0051] Regarding step S2 of this protocol:

[0052] To scientifically quantify the impact of secondary indicators on primary indicators, this solution employs an objective weighting strategy that couples the entropy weight method with the coefficient of variation method. The entropy weight method quantifies the degree of dispersion of indicator data through information entropy, thereby mining the information contained in the indicator. The coefficient of variation method dynamically assesses the importance of the indicator based on its degree of variation. This organic combination of the two complements the advantages of objective weighting methods and forms a more robust secondary indicator weight matrix.

[0053] In step S2, a coupled data set of the magnetorheological damper at different frequencies and currents is obtained through an orthogonally designed bench test, and the original sample data values of the secondary indicators are obtained based on the coupled data set, where the coupled data set includes displacement, velocity, and damping force response signals.

[0054] Specifically, in u Frequency and v Under the working condition combination of the current, the displacement, velocity and damping force response signals of the magnetorheological damper are collected synchronously as a coupling data set, and the coupling data set is input into the mechanical model of the magnetorheological damper to be evaluated for parameter identification. The results of the first step of the iterative optimization are obtained for each working condition combination. f Frequency (1≤ f ≤ u ) and I Current (0≤ I ≤ v ) under the working condition combination of the secondary index u ij The original sample data value x ij-f-I .

[0055] In view of the strong nonlinearity, multi-parameter coupling and uncertainty characteristics of magnetorheological damper modeling, in order to eliminate the influence of the differences in the dimensions and numerical ranges of different indicators on the objective weight calculation results, step S2 of this solution further includes the following steps:

[0056] The original sample data values are standardized to obtain standardized data values. The entropy method is used to calculate the standardized sample data values to obtain the entropy weight method weight of each secondary indicator. The coefficient of variation method is used to calculate the standardized sample data values to obtain the coefficient of variation method weight of each secondary indicator. The weighted entropy weight method weight and the coefficient of variation method weight are added to obtain the objective weight of each secondary indicator. The objective weights of the secondary indicators under each primary indicator are summarized to obtain the secondary indicator weight matrix.

[0057] This scheme is for the original sample data value x ij-f-I Perform standardization to obtain standardized sample data values , when the original sample data value is a positive indicator, the formula for standardization is: ;

[0058] When the original sample data value is a negative indicator, the formula for standardization is: ;

[0059] Where Min(x ij-f-I ) is x ij-f-I The minimum value, Max(x ij-f-I ) is x ij-f-I The maximum value of .

[0060] This scheme uses the entropy method to calculate the standardized sample data value to obtain the entropy weight method weight of each secondary indicator as follows:

[0061] For each secondary indicator, obtain the proportion of each standardized sample data value corresponding to the secondary indicator in the sum of the standardized sample data values of all working condition combinations, calculate the entropy value of the secondary indicator based on the proportion, and calculate the entropy weight method weight of the secondary indicator under the corresponding primary indicator based on the entropy value of the secondary indicator.

[0062] The specific calculation formula is as follows:

[0063] ;

[0064] ;

[0065] ;

[0066] where p ij-f-I is the proportion of each standardized sample data value in the sum of the standardized sample data values of all working condition combinations, is the standardized sample data value, f is the frequency, u is the total number of frequencies of the operating condition combinations, where 1≤ f ≤ u , I is the current, v is the total current of the working condition combination, where 0≤I ≤ v , b ij is the entropy value of the jth secondary indicator under the i-th primary indicator, ln() is the logarithm, w ije is the entropy weight of the jth secondary indicator under the ith first-level indicator under the corresponding i-th first-level indicator, where n is the total number of secondary indicators under the first-level indicator.

[0067] This scheme uses the coefficient of variation method to calculate the standardized sample data value to obtain the coefficient of variation weight of each secondary indicator as follows:

[0068] Take the mean and standard deviation of the standardized sample data values corresponding to the secondary indicator under all working condition combinations, use the quotient of the standard deviation and the mean as the coefficient of variation of the secondary indicator, and calculate the coefficient of variation method weight of the secondary indicator based on the coefficient of variation.

[0069] The specific calculation formula is as follows:

[0070] ;

[0071] ;

[0072] ;

[0073] ;

[0074] in is the mean, is the standard deviation, is the coefficient of variation, w ijc is the weight of the coefficient of variation method, is the standardized sample data value, f is the frequency, u is the total number of frequencies of the operating condition combinations, where 1≤ f ≤ u , I is the current, v is the total current of the working condition combination, where 0≤ I ≤ v , n is the total number of secondary indicators under the primary indicator.

[0075] This scheme uses the entropy weight method to quantify the degree of indicator dispersion through information entropy theory, determines the relative importance of each secondary indicator in the comprehensive evaluation process, and measures the degree of dispersion of secondary indicators based on the weight of the coefficient of variation method, so as to compare the volatility and variability of different secondary indicators. After obtaining the entropy weight method weight and the coefficient of variation method weight, this scheme weights the entropy weight method weight and the coefficient of variation method weight to obtain the objective weight of each secondary indicator, and normalizes the secondary indicators of the same primary indicator and summarizes them to obtain the secondary indicator weight matrix. The specific calculation is as follows:

[0076] ;

[0077] ;

[0078] in is the preference coefficient, which can be 0.5 according to experience. w ije is the entropy weight method weight, w ijc is the weight of the coefficient of variation method, W i is the weight matrix of the secondary indicators under the i-th primary indicator, w ij is the objective weight of the jth secondary indicator under the ith first-level indicator, and n is the total number of secondary indicators under the ith first-level indicator.

[0079] The formula for normalization is: .

[0080] Regarding step S3 of this protocol:

[0081] In order to quantify the multidimensional fuzzy characteristics of the dynamic performance of magnetorheological dampers, this proposal constructs a quantitative model based on the fuzzy comprehensive evaluation method. In order to address the problem of fuzzy indicator boundaries caused by the inherent time-varying nonlinear characteristics of magnetorheological dampers, the membership function is used to convert the fuzzy performance evaluation concepts into quantitative values, thereby achieving accurate mapping of the multidimensional performance of the mechanical model.

[0082] Specifically, step S3 of "dividing the levels based on the global interval of the original sample data values of each secondary indicator to obtain different evaluation levels, and converting the evaluation level of each secondary indicator into a secondary indicator score vector" further includes the steps of: obtaining the global interval of the original sample data values of each secondary indicator, dividing the global interval into multiple mutually exclusive and complete secondary indicator evaluation intervals, each secondary indicator evaluation interval is defined as a unique evaluation level, and converting the evaluation level into a scoring vector to obtain the secondary indicator scoring vector of the current secondary indicator.

[0083] In some embodiments, the original domain interval of the secondary indicators under the same primary indicator is divided into the same number of secondary indicator evaluation intervals, and the number of secondary indicator evaluation intervals is controlled between 3 and 7.

[0084] Specifically, obtain each secondary indicator u ij In the f Frequency (1≤ f ≤ u ) and I Current (0≤ I ≤ v ) under the working condition combination, and divide the global interval into several mutually exclusive and complete secondary index evaluation intervals I according to the empirical method. d (1≤ d ≤ b , b is the total number of evaluation intervals of the secondary indicators) to balance the contradiction between discrimination and fuzziness.

[0085] A unique evaluation level is defined for each secondary indicator evaluation interval, as follows:

[0086] ;

[0087] where e1, e2...e b is the evaluation grade corresponding to the secondary indicator evaluation interval, b is the total number of secondary indicator evaluation intervals), and E is the grade sequence.

[0088] Furthermore, in order to achieve numerical evaluation, this solution further converts the evaluation level into a scoring vector to obtain the secondary indicator scoring vector V of the current secondary indicator, which is expressed as follows: ;

[0089] in v 1, v 2... v b is the score of the secondary indicator evaluation level, b is the total number of secondary indicator evaluation intervals, and V is the secondary indicator scoring vector, where v b ∈[0,100] and arranged in descending order according to an arithmetic progression.

[0090] In some embodiments, when b=4, it can be defined as a grade sequence of {excellent, good, medium, poor}, then define .

[0091] Furthermore, in order to describe the fuzzy correlation and interaction between different secondary indicators, this solution uses the membership calculation method to obtain the fuzzy relationship matrix of the secondary indicators. Specifically, the ratio of the number of original sample data values of each secondary indicator that meets the evaluation level to the total number of original sample data values of the secondary indicator is used as the membership of the secondary indicator under the evaluation level. This process is essentially a nonlinear mapping relationship that transforms the numerical space into the fuzzy evaluation space. The calculation formula is as follows:

[0092] ;

[0093] in a d is the original sample data value of the jth secondary indicator under the i-th primary indicator Satisfy the d-th secondary index evaluation interval I d The number of samples, uv ij is the total number of original sample data values of the jth secondary indicator under the i-th primary indicator, is the membership degree of the jth secondary indicator under the ith first-level indicator at the dth evaluation level.

[0094] The membership matrix is obtained by arranging the membership of each secondary indicator under different evaluation levels, and the fuzzy relationship matrix of the secondary indicators under the same primary indicator is obtained by splicing the membership matrices of the secondary indicators, which can be expressed as:

[0095] ;

[0096] in R i is the fuzzy relationship matrix of the secondary indicators of the i-th first-level indicator. The fuzzy relationship matrix of the secondary indicators is a matrix with n rows and b columns, where n is the number of secondary indicators under the i-th first-level indicator, and b is the number of evaluation levels of the secondary indicators. is the membership degree of the nth secondary indicator under the i-th primary indicator at the b-th evaluation level.

[0097] Furthermore, the secondary indicator fuzzy comprehensive evaluation matrix is obtained by weighting the secondary indicator fuzzy relationship matrix according to the secondary indicator weight matrix. The calculation formula is as follows:

[0098] ;

[0099] where Q i is the fuzzy comprehensive evaluation matrix of the second-level indicators of the i-th first-level indicator, W i is the secondary indicator weight matrix of the i-th primary indicator, where R iis the fuzzy relationship matrix of the second-level indicators of the i-th first-level indicator.

[0100] Furthermore, the scalar product operation is performed on the secondary indicator fuzzy relationship matrix and the secondary indicator score vector to obtain the comprehensive score of the primary indicator. The calculation formula is as follows:

[0101] ;

[0102] where Q i is the fuzzy comprehensive evaluation matrix of the second-level indicators of the i-th first-level indicator, V is the second-level indicator scoring vector, z i is the comprehensive score of the first-level indicators of the i-th first-level indicator.

[0103] Regarding step S4 of this protocol:

[0104] To construct a multidimensional evaluation system for the mechanical model of magnetorheological dampers, this solution uses the Analytic Hierarchy Process (AHP) to determine the subjective weights of primary indicators. For the three primary indicators—model fidelity, model business value, and model software-driven quality—a judgment matrix is constructed based on expert experience to transform qualitative perceptions into quantitative weights.

[0105] Specifically, step S4 further includes the steps of:

[0106] Obtain the subjective evaluation of the experts on the relative importance of each first-level indicator, and construct the first-level indicator judgment matrix of the first-level indicator based on the subjective evaluation of relative importance. Calculate the geometric mean of the elements in each row of the first-level indicator judgment matrix and perform normalization to obtain the first-level indicator weight matrix, where the subjective weight of each first-level indicator is recorded in the first-level indicator weight matrix.

[0107] In some embodiments, a relative importance subjective evaluation is obtained, and a primary indicator judgment matrix of the primary indicator is constructed based on the relative importance subjective evaluation:

[0108] ;

[0109] where h pq It is the relative quantitative index of the pth (p=1,2,3) first-level index relative to the qth (q=1,2,3) first-level index.

[0110] This scheme obtains the subjective evaluation of relative importance according to the quantitative scale conversion table shown in Table 1:

[0111] Table 1 Quantization scale conversion table

[0112]

[0113] The formula for calculating the first-level indicator weight matrix is as follows:

[0114] ;

[0115] Where r is the order of the first-level indicator judgment matrix A, r=3; h pq is the relative quantitative index of the pth (p=1,2,3) first-level index relative to the qth (q=1,2,3) first-level index, It is the continuous multiplication symbol, and the elements with subscript q from 1 to r are multiplied continuously.

[0116] In some embodiments, the consistency verification of the first-level indicator judgment matrix is performed. If the consistency verification fails, the first-level indicator judgment matrix is adjusted.

[0117] About consistency verification:

[0118] Calculate the maximum eigenvalue of the first-level indicator judgment matrix, and calculate the consistency ratio based on the maximum eigenvalue. When the consistency ratio is less than the set threshold, it is considered that the current first-level indicator judgment matrix passes the consistency judgment. The calculation formula is as follows:

[0119] ;

[0120] ;

[0121] Where CR is the consistency ratio, RI () is the random consistency index of the first-level indicator judgment matrix A, RI = 0.52, r is the order of the first-level indicator judgment matrix A, r = 3, is the maximum eigenvalue.

[0122] In some embodiments, when the consistency ratio CR is less than 0.1, the first-level index judgment matrix passes the consistency verification.

[0123] After knowing the subjective weight of the primary index and the objective weight of the secondary index, in order to achieve the quantification and objective reflection of the internal relationship between the primary and secondary indicators, this scheme organically integrates the objective weight of the secondary index, the comprehensive score of the primary index and the subjective weight of the primary index, and constructs the comprehensive performance index J of the mechanical model of the magnetorheological vibration damper as the quantitative score of the mechanical model of the magnetorheological vibration damper, which is expressed as:

[0124] ;

[0125] ;

[0126] ;

[0127] where w i For U i The corresponding subjective weight of the first-level indicator, z i For U iThe corresponding first-level indicator comprehensive score, U i is the i-th first-level indicator, W i is the secondary indicator weight matrix of the i-th secondary indicator, R i is the secondary indicator simulation relationship matrix of the i-th secondary indicator, V is the scoring vector of the secondary indicator score; where w 11 、w 21 for u 11 ,u 12 The objective weight of the corresponding secondary indicator, w 21 、w 22 、w 23 for u 21 ,u 22， u 23 The objective weight of the corresponding secondary indicator; w 31 、w 32 for u 31 ,u 32 The objective weight of the corresponding secondary indicator, R i is the fuzzy relationship matrix of the second-level indicators of the i-th first-level indicator.

[0128] Example 2

[0129] This second embodiment will provide an embodiment of the magnetorheological vibration damper mechanical model evaluation method based on the fuzzy comprehensive evaluation method when actually evaluating the mechanical model of the magnetorheological vibration damper.

[0130] First, a bench test of the magnetorheological damper was carried out, and the bench test loading frequency was set to a sine wave of 1-4 Hz with a step size of 1 Hz; the control current range was set to 0-3 A with a step size of 0.5 A, and the corresponding setting frequency f working condition number u value was 4, and the current I working condition number v value was 7.

[0131] The damping force test data under different working condition combinations were collected, with 206 data points collected under each working condition. Therefore, the t value of the number of model damping force samples used to calculate the root mean square error and determination coefficient of the model damping force is 206.

[0132] To verify the effectiveness of the evaluation method proposed in this paper, a comparative analysis was conducted using the Bingham model and the Magic Formula model of a magnetorheological damper using case studies. Based on bench test data from the magnetorheological damper, both models were simulated using MATLAB / Simulink to obtain data for each set of secondary indicators under different frequencies and current conditions. These included the root mean square error (RMSE) and coefficient of determination (R²) of the model damping force, the relative rate of change of the total area of the dynamometer diagram, the relative rate of change of the area ratio of the dynamometer diagram, and the relative rate of change of the saturation of the dynamometer diagram, as well as the model runtime and memory usage. Based on this data, the evaluation method proposed in this paper calculated the comprehensive scores of the three primary indicators of the two models, as well as the final quantitative score of the magnetorheological damper mechanical model. The differences were then compared. Through this comparative analysis, the advantages and disadvantages of the two models in terms of accuracy, stability, and computational efficiency were systematically evaluated.

[0133] Example 1 - Bingham Model

[0134] First, the secondary index data corresponding to the primary index are obtained based on the simulation results of the Bingham model. The root mean square error (RMSE) and the coefficient of determination (R) of the damping force corresponding to the model fidelity under different frequency and current conditions are obtained. 2 See Table 2:

[0135] Table 2 Model assurance results of Bingham model

[0136] .

[0137] Combined with the displacement and damping force values fitted by the Bingham model, the shoelace formula in Excel was used to calculate the fitted data. The relative change rate of the total area of the dynamometer diagram, the relative change rate of the area ratio of the dynamometer diagram, and the relative change rate of the saturation of the dynamometer diagram corresponding to the model business value under different frequencies and different current conditions were obtained, as shown in Table 3:

[0138] Table 3. Model business value results of Bingham model

[0139] .

[0140] The model fitting running time and memory usage data of the Bingham model under different frequency and current conditions are summarized in Table 4:

[0141] Table 4 Bingham model software driver quality results

[0142] .

[0143] Based on the aforementioned magnetorheological damper mechanical model evaluation method, the calculation of the Bingham model is as follows:

[0144] I. Calculate the objective weights of the secondary indicators of the Bingham model:

[0145] The fidelity results of the Bingham model in Table 2 are normalized, and the results are shown in Table 5.

[0146] Table 5 Standardized processing results of Bingham model fidelity

[0147] .

[0148] The entropy weight method is used to calculate the entropy weight of the secondary indicators. The corresponding proportions are shown in Table 6:

[0149] Table 6. Bingham model fidelity ratio results

[0150] .

[0151] Combined with Table 6, the RMSE value and R 2 The entropy value corresponding to the value b 11 =0.9460, b 12 =0.9500, and then use the entropy value to calculate the RMSE value and R 2 Entropy weight of the value w 11e =0.522, w 12e =0.478, based on the RMSE value and R calculated in Table 5 2 The mean corresponding to the value and standard deviation , and thus calculate the RMSE value and R 2 The coefficient of variation corresponding to the value , calculate the RMSE value and R 2 Coefficient of variation weights w 11c =0.498, w 12c =0.502, combining the entropy weight method weight and the coefficient of variation method weight calculated above to comprehensively calculate the RMSE and R² secondary indicator objective weights of the Bingham model w 11 and w 12 , forming the secondary indicator weight matrix:

[0152] ;

[0153] Similarly, calculate the secondary indicator weight matrix corresponding to the model business value and model software driving quality of the Bingham model W 2, W 3 are as follows:

[0154]

[0155] .

[0156] II. Calculate the comprehensive score of the first-level indicators:

[0157] According to the data distribution characteristics of RMSE and R² of the Bingham model fitting, the data is divided into four secondary indicator evaluation intervals based on the empirical method, and the corresponding four secondary indicator evaluation levels constitute the evaluation level set:

[0158] E ={excellent, good, average, poor};

[0159] Evaluation level corresponding to the secondary indicator score vector V as follows:

[0160] ;

[0161] The secondary indicator evaluation intervals, secondary indicator evaluation levels, and score relationships for model fidelity are shown in Table 7:

[0162] Table 7 Evaluation intervals of secondary indicators of model fidelity

[0163] .

[0164] Combining the values in Table 2 with the secondary index division intervals in Table 7, the secondary index fuzzy relationship matrix of the Bingham model fidelity is calculated. R 1 is as follows:

[0165] ;

[0166] Reuse R 1 Combined with the secondary indicator weight vector W 1 Calculate the fuzzy comprehensive evaluation matrix of the secondary indicators:

[0167] ;

[0168] Combined calculation of the first-level indicator comprehensive score of the Bingham model fidelity z 1 is as follows:

[0169] .

[0170] Since the aforementioned secondary indicator evaluation scores are all based on a percentage evaluation method, the same method is used for the comprehensive score of the first-level indicators of the Bingham model. The comprehensive score of the first-level indicators of the Bingham model fidelity is 65.3232, which is between 55 and 70, and the corresponding Bingham model fidelity evaluation level is between poor and medium.

[0171] The same secondary indicator evaluation level and corresponding secondary indicator score vector are used as for model fidelity. Based on the data distribution characteristics and empirical method, the secondary indicator evaluation interval, secondary indicator evaluation level and score relationship of the model business value are divided as shown in Table 8:

[0172] Table 8 Evaluation range of secondary indicators of model business value

[0173] .

[0174] Combine the values in Table 3 with the secondary indicator evaluation intervals in Table 8 to calculate the secondary indicator fuzzy relationship matrix of the Bingham model business value. R 2 is as follows:

[0175] ;

[0176] Calculate the secondary indicator fuzzy comprehensive evaluation matrix of Bingham model business value

[0177] ;

[0178] Finally, calculate the comprehensive score of the first-level indicators of the Bingham model business value z 2 is as follows:

[0179] ;

[0180] The comprehensive score of the first-level indicator of business value of the Bingham model is 69.9154, which is between 55 and 70, and the corresponding Bingham model business value evaluation level is between poor and medium.

[0181] The same secondary indicator evaluation levels and corresponding secondary indicator score vectors are used as above. According to the data distribution characteristics, the secondary indicator evaluation intervals, secondary indicator evaluation levels and score relationships of the model software driver quality are divided according to the empirical method. See Table 9.

[0182] Table 9 Evaluation intervals of secondary indicators of model software driver quality

[0183] .

[0184] Combine the values in Table 4 with the secondary indicator evaluation intervals in Table 9 to calculate the secondary indicator fuzzy relationship matrix of the Bingham model software driver quality.R 3 are as follows:

[0185] ;

[0186] Calculate the fuzzy comprehensive evaluation matrix of secondary indicators:

[0187] ;

[0188] Finally, calculate the comprehensive score of the first-level indicators of the Bingham model software driver quality z 3 are as follows:

[0189] ;

[0190] The comprehensive score of the first-level indicator of software driver quality of the Bingham model is 93.1, which is between 85 and 100, and the corresponding Bingham model software driver quality evaluation level is between good and excellent.

[0191] III. Calculation of subjective weights and quantitative indicators of primary indicators

[0192] Using the hierarchical analysis method, we consider the three first-level indicators of model fidelity, model business value, and model software-driven quality, and compare the relative importance of each indicator. The comparison results are shown in Table 10.

[0193] Table 11 Correspondence table of the importance of indicators

[0194] .

[0195] The first-level evaluation index judgment matrix A is established based on Table 10 as follows:

[0196] ;

[0197] Calculate the subjective weight of the first-level indicator w i as follows:

[0198] ;

[0199] Calculate the maximum eigenvalue of the first-level evaluation index judgment matrix and consistency ratio , the consistency check passed.

[0200] Combined calculation of the first-level index comprehensive score z 1, z 2, z 3. Subjective weight of the first-level indicator calculated above w 1, w 2, w 3. Finally calculate the comprehensive performance index of the Bingham modelJ as follows:

[0201] ;

[0202] Based on the aforementioned percentage-based evaluation method, the same method was used for the model's comprehensive performance index. The Bingham model's comprehensive performance index was 68.4292, ranging from 55 to 70, corresponding to a Bingham model comprehensive evaluation grade between poor and medium.

[0203] Example 2

[0204] First, the secondary indicator data corresponding to the primary indicator is obtained based on the simulation results of the magic formula model. The root mean square error RMSE and determination coefficient R of the model damping force corresponding to the model fidelity under different frequency and current working conditions are: 2 See Table 11:

[0205] Table 11 Model assurance results of the magic formula model

[0206] .

[0207] Combining the displacement and damping force values fitted by the magic formula model, the shoelace formula in Excel is used to calculate the fitted data. The relative change rate of the total area of the dynamometer diagram, the relative change rate of the area ratio of the dynamometer diagram, and the relative change rate of the saturation of the dynamometer diagram corresponding to the model business value under different frequencies and different current conditions are shown in Table 12:

[0208] Table 12: Bingham model business value results

[0209] .

[0210] Table 13 summarizes the model fitting running time and memory usage data of the magic formula model under different frequency and current conditions:

[0211] Table 13 Magic Model Software Driver Quality Results

[0212] .

[0213] Based on the aforementioned magnetorheological damper mechanical model evaluation method, the calculation of the magic formula model is as follows:

[0214] I. Calculate the objective weights of the secondary indicators of the magic formula model:

[0215] The magic formula model fidelity results in Table 11 are normalized, and the results are shown in Table 14.

[0216] Table 14. Standardized processing results of magic formula model fidelity

[0217] .

[0218] The entropy weight method is used to calculate the entropy weight of the secondary indicators. The corresponding proportions are shown in Table 15:

[0219] Table 15 The proportion of magic formula model fidelity

[0220] .

[0221] Combined with Table 15, the RMSE value and R 2 The entropy value corresponding to the value b 11 =0.9600, b 12 =0.9530, and then use the entropy value to calculate the RMSE value and R 2 Entropy weight of the value w 11e =0.456, w 12e =0.544, and the RMSE value and R 2 The mean corresponding to the value and standard deviation , calculate the RMSE value and R 2 The coefficient of variation corresponding to the value , calculate the RMSE value and R 2 Coefficient of variation weights w 11c =0.488, w 12c =0.512, combining the entropy weight method weight and the coefficient of variation method weight calculated above to comprehensively calculate the RMSE and R² secondary evaluation index objective weights of the magic formula model w 11 and w 12 , forming the secondary evaluation index weight matrix

[0222] ;

[0223] Calculate the secondary evaluation indicator weight vector corresponding to the model business value and model software driving quality of the magic formula model W 2, W 3 are as follows:

[0224]

[0225] .

[0226] II. Calculate the comprehensive score of the first-level indicators:

[0227] According to the data distribution characteristics of RMSE and R² of the magic formula model fitting, the data is divided into four secondary indicator evaluation intervals based on the empirical method, and the corresponding four secondary indicator evaluation levels form an evaluation level set:

[0228] E ={Excellent, Good, Average, Poor}

[0229] Evaluation level corresponding to the secondary indicator score vector V as follows:

[0230] ;

[0231] Combine the values in Table 11 with the secondary index division intervals in Table 7 to calculate the secondary index fuzzy relationship matrix of the magic formula model fidelity R 1 is as follows:

[0232] ;

[0233] Reuse R 1 Combined with the weight vector of the secondary evaluation index W 1 Calculate the fuzzy comprehensive evaluation matrix of secondary evaluation indicators

[0234] ;

[0235] Calculate the comprehensive score of the first-level evaluation index of the magic formula model fidelity z 1 is as follows:

[0236] .

[0237] The comprehensive score of the first-level evaluation index of the Magic Formula model fidelity is 85.8786, which is between 85 and 100, and the corresponding Bingham model fidelity evaluation level is between good and excellent.

[0238] Calculate the fuzzy relationship matrix of the secondary evaluation indicators of the magic formula model business value R 2 is as follows:

[0239] ;

[0240] Calculate the secondary evaluation index fuzzy comprehensive evaluation matrix of the magic formula model business value

[0241] ;

[0242] Finally, calculate the comprehensive score of the first-level evaluation indicators of the magic formula model's business value z 2 is as follows:

[0243] ;

[0244] The comprehensive score of the first-level evaluation index of the Magic Formula Model's business value is 83.8879, which is between 70 and 85, and the corresponding Magic Formula Model's business value evaluation level is between medium and good.

[0245] Combine the values in Table 13 with the intervals divided in Table 9 to calculate the fuzzy relationship matrix of the secondary evaluation indicators of software driver quality of the magic formula model. R 3 are as follows:

[0246] ;

[0247] Calculate the fuzzy comprehensive evaluation matrix of secondary evaluation indicators

[0248] ;

[0249] Finally, calculate the comprehensive score of the first-level evaluation indicators of the magic formula model software driver quality z 3 are as follows:

[0250] ;

[0251] The comprehensive score of the first-level evaluation index of the magic formula model software driver quality is 64.8871, which is between 55 and 70. The corresponding magic formula model software driver quality evaluation level is between poor and medium.

[0252] III. Calculation of subjective weights and quantitative indicators of primary indicators

[0253] Calculate the subjective weight of the first-level indicator w i as follows:

[0254] ;

[0255] Finally, calculate the comprehensive performance index of the Bingham model J as follows:

[0256] ;

[0257] Based on the aforementioned percentage evaluation method, the comprehensive performance evaluation index of the magic formula model is 83.8518, which is between 70 and 85, and the corresponding comprehensive evaluation level of the Bingham model is between medium and good.

[0258] In summary, according to the fuzzy comprehensive evaluation method process, the comprehensive performance evaluation index of the Bingham model is 68.4292, while the comprehensive performance evaluation index of the magic formula is 83.8518. It can be concluded that the comprehensive performance of the magic formula model is better than that of the Bingham model. Specifically, in terms of model fidelity, the Bingham model scored 65.3232, and the Magic Formula model scored 85.8786, which was 31.47% higher than the Bingham model. This significant difference indicates that the Magic Formula model is significantly better than the Bingham model in damping force prediction accuracy. In terms of model business value, the Bingham model scored 69.9154, and the Magic Formula model scored 83.8879, which was 19.98% higher than the Bingham model, indicating that the Magic Formula model is better than the Bingham model in terms of hysteresis curve shape. In terms of model software-driven quality, the Bingham model scored 91.3257, and the Magic Formula model scored 64.8871, which was 40.75% higher than the Magic Formula model, reflecting that the Bingham model is better than the Magic Formula model in terms of computational efficiency and memory consumption.

[0259] Those skilled in the art should understand that the technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0260] The above embodiments merely illustrate several embodiments of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method, characterized in that: The following steps are involved: S1: Determine the indicators: The model fidelity, business value, and software-driven quality of the magnetorheological damper are used as primary indicators. The relative rate of change of the total area of the model dynamometer, the relative rate of change of the area ratio of the dynamometer, and the relative rate of change of the saturation of the dynamometer are used as secondary indicators of the business value of the model. The model running time and model memory usage are used as secondary indicators of the software-driven quality of the model. The root mean square error and coefficient of determination between the predicted value and the experimental value of the mechanical model damping force are used as secondary indicators of the model fidelity. S2: Determine the objective weight of the secondary indicators; Obtain the original sample data values of the secondary indicators of the magnetorheological shock absorber under different frequencies and currents, standardize the original sample data values to obtain standardized data values, use the entropy method to calculate the standardized sample data values to obtain the entropy weight method weight of each secondary indicator, use the coefficient of variation method to calculate the standardized sample data values to obtain the coefficient of variation method weight of each secondary indicator, add the weighted entropy weight method weight and the coefficient of variation method weight to obtain the objective weight of each secondary indicator, and summarize the objective weights of the secondary indicators under each primary indicator to obtain the secondary indicator weight matrix; S3: Calculate the comprehensive score of the first-level indicators based on the objective weights of the second-level indicators and the original sample data values: Based on the global interval of the original sample data values of each secondary indicator, different evaluation levels are obtained, the evaluation level of each secondary indicator is converted into a secondary indicator score vector, the ratio of the number of original sample data values of each secondary indicator that meets the evaluation level to the total number of original sample data values of the secondary indicator is used as the membership of the secondary indicator at the evaluation level, the membership of each secondary indicator at different evaluation levels is sorted to obtain a membership matrix, the membership matrices of the secondary indicators under the same primary indicator are concatenated to obtain a secondary indicator fuzzy relationship matrix, the secondary indicator fuzzy relationship matrix is weighted according to the secondary indicator weight matrix to obtain a secondary indicator fuzzy comprehensive evaluation matrix, and a scalar product operation is performed on the secondary indicator fuzzy relationship matrix and the secondary indicator score vector to obtain a comprehensive score of the primary indicator; S4: Determine the subjective weight of the first-level indicators based on the hierarchical analysis method, and obtain the quantitative score of the magnetorheological shock absorber mechanical model based on the comprehensive score of the first-level indicators weighted by the subjective weight.

2. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1 is characterized in that: For each secondary indicator, obtain the proportion of each standardized sample data value corresponding to the secondary indicator in the sum of the standardized sample data values of all working condition combinations, calculate the entropy value of the secondary indicator based on the proportion, and calculate the entropy weight method weight of the secondary indicator under the corresponding primary indicator based on the entropy value of the secondary indicator.

3. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1 is characterized in that: Take the mean and standard deviation of the standardized sample data values corresponding to the secondary indicator under all working condition combinations, use the quotient of the standard deviation and the mean as the coefficient of variation of the secondary indicator, and calculate the coefficient of variation method weight of the secondary indicator based on the coefficient of variation.

4. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1 is characterized in that: The global interval of the original sample data value of each secondary indicator is obtained, and the global interval is divided into multiple mutually exclusive and complete secondary indicator evaluation intervals. Each secondary indicator evaluation interval is defined as a unique evaluation level, and the evaluation level is converted into a scoring vector to obtain the secondary indicator scoring vector of the current secondary indicator.

5. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1, characterized in that: The formula for calculating the membership degree of the secondary indicator at this evaluation level is as follows: ； in a d is the original sample data value of the jth secondary indicator under the i-th primary indicator Satisfy the d-th secondary index evaluation interval I d The number of samples, uv ij is the total number of original sample data values of the jth secondary indicator under the i-th primary indicator, is the membership degree of the jth secondary indicator under the ith first-level indicator at the dth evaluation level; The membership matrix is obtained by arranging the membership of each secondary indicator under different evaluation levels, and the fuzzy relationship matrix of the secondary indicators under the same primary indicator is obtained by splicing the membership matrix of the secondary indicators: ； in R i is the fuzzy relationship matrix of the secondary indicators of the i-th first-level indicator. The fuzzy relationship matrix of the secondary indicators is a matrix with n rows and b columns, where n is the number of secondary indicators under the i-th first-level indicator, and b is the number of evaluation levels of the secondary indicators. is the membership degree of the nth secondary indicator under the i-th primary indicator at the b-th evaluation level.

6. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1, characterized in that: Obtain the subjective evaluation of the experts on the relative importance of each first-level indicator, and construct the first-level indicator judgment matrix of the first-level indicator based on the subjective evaluation of relative importance. Calculate the geometric mean of the elements in each row of the first-level indicator judgment matrix and perform normalization to obtain the first-level indicator weight matrix, where the subjective weight of each first-level indicator is recorded in the first-level indicator weight matrix.

7. The magnetorheological damper mechanical model evaluation method based on fuzzy comprehensive evaluation method according to claim 1, characterized in that: The original domain interval of the secondary indicators under the same primary indicator is divided into the same number of secondary indicator evaluation intervals, and the number of secondary indicator evaluation intervals is controlled between 3 and 7.

Images