A method and device for active suspension parameter dynamic design under digital mapping

Abstract

The application discloses a kind of active suspension parameter dynamic design method and device under digital mapping, the method includes: the parameter dataset x of digital virtual active suspension system is constructed;The road environment of digital virtual active suspension system is constructed;The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in its dataset y are constructed, determine the number of the output state dataset y of the digital virtual active suspension system is grabbed wheel, then the output state dataset y of the digital virtual active suspension system is reconstructed, obtains new simulation output state datasetBased on multidimensional mixed index, establish objective function, based on the objective function, the numerical value of each parameter in the output state dataset y of the digital virtual active suspension system is feedback optimized.This method can overcome the problems of difficult prediction, high cost, low reliability and unstable safety in traditional solutions.

Description

Technical Field

[0001] This invention relates to the field of vehicle suspension design, and more specifically to a method and apparatus for dynamic design of active suspension parameters under digital mapping. Background Technology

[0002] As a crucial component of the vehicle chassis system, the suspension system plays a vital role in damping vibrations and ensuring occupant safety. Simultaneously, changes in the suspension system parameters affect various aspects of the vehicle's performance, including ride comfort, handling, stability, driver comfort, and the integrity of the cargo carried.

[0003] Based on whether or not a power source is present in the suspension system, current suspension systems can be divided into passive suspension, semi-active suspension, and active suspension. Passive suspension lacks an internal power source and its parameters are not adjustable, resulting in poor adaptability to different environments in practical applications. Semi-active suspension has partially adjustable parameters, enabling real-time dynamic adjustment to improve and optimize vehicle performance under various conditions. Active suspension, with its own power source, achieves better control over vehicle pitch, roll, and yaw movements, which is of great significance and value for improving overall vehicle performance, thus research on it has gradually become a hot topic.

[0004] An active suspension system mainly consists of three parts: springs, dampers, and actuators. Figure 1 As shown, the corresponding main design parameters are: suspension spring stiffness coefficient K, suspension damping coefficient C, and actuator electromagnetic thrust coefficient k. F Based on this, it is necessary to further consider the nonlinear coefficient ξ of the spring when it operates in the nonlinear range, and the body roll center height H of the vehicle after the suspension is installed. roll .

[0005] Suspension parameters should match the vehicle's parameters. Traditional suspension parameter design relies on multiple stages, including system modeling, parameter tuning, bench testing, and vehicle verification. This makes the parameter design and optimization process for suspension systems cumbersome, and can lead to inconsistencies in product performance. Furthermore, traditional suspension system parameters cannot be adjusted for different real-time road conditions and complex environments. Once system parameters are set, system performance is also determined, which reduces the tolerance for errors in the product design process and increases the demands on designers. Therefore, intelligent online real-time optimization design of suspension under real-time road conditions has become a new product requirement and a critical issue that urgently needs to be addressed.

[0006] This invention proposes a dynamic design method for active suspension parameters under digital mapping. This method randomly pre-sets suspension parameter values ​​in a digital system, reconstructs the dynamic response data of the suspension and vehicle system, extracts features from the suspension parameters, and finally optimizes the suspension system parameters by outputting the state, thereby achieving a comprehensive optimization design of the pre-performance of the suspension system parameters. Summary of the Invention

[0007] In view of this, the present invention provides a method and apparatus for dynamic design of active suspension parameters under digital mapping, which can solve the technical problem of dynamic design and real-time iterative optimization of suspension parameters in the process of multi-dimensional data fusion under virtual data mapping conditions.

[0008] To solve the above-mentioned technical problems, the present invention is implemented as follows.

[0009] A method for dynamic design of active suspension parameters under digital mapping includes: Step S1: Set the suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H roll Initial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Step S2: Obtain the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Step S3: For each Perform feature extraction and calculate the extracted feature and entropy matrix EN. iThe distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function.

[0010] Preferably, step S1 includes: Step S11: Based on the current load of the vehicle and the requirements of the design objectives, set the suspension stiffness coefficient K, suspension damping coefficient C, electromagnetic thrust coefficient kF, suspension nonlinearity coefficient ξ, and vehicle roll center height H. roll The initial values ​​are used to form the parameter dataset x=[x1, x2, x3, x4, x5] of the digital virtual active suspension system. T =[K, C, k F , ξ, H roll ] T ; Step S12: Calculate the road adhesion coefficient μ and the road surface roughness coefficient G. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r As external environmental data, a road environment is constructed for the digital virtual active suspension system; the variation range of the road adhesion coefficient μ is set to (0, 1), the variation range of the vehicle speed V is set to [0, 150km / h], and the road roughness coefficient G... q The range of variation is set to [16, 16384].

[0011] Preferably, step S2 includes: Step S21: Based on the parameter dataset x of the digital virtual active suspension system and the road environment of the digital virtual active suspension system, determine the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T As state parameters representing the output state of the digital virtual active suspension system; obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T Construct the output state dataset y of the digital virtual active suspension system. y =[ y 1, y 2, y 3, y 4, y 5] T =[ D , , , v s , F T T ; Among them, under the vehicle stable state, the output variable y i corresponds to a safety threshold range H i ={[y imin , y imax}, and the output variable y i is a discrete, multi-dimensional data, and different dimensions represent the system sampling data at different moments, ; Step S22: Set the number of data capture rounds TotalNum , initialize the number of data capture rounds knum to 0, and set TotalNum equal to 1000T, where T is the current time; Step S23: For each output variable y i : Based on the output variable y i corresponding to the safety threshold range H i ={[y imin ,y imax} and the output variable y i , determine the effective capture times n of the output state data set of the digital virtual active suspension system, and n <1000 T ;

[0012] Among them, is the number of data in y i that exceeds the safety threshold range H i , is the ceiling function; Randomly capture the data in each output variable y i n times to determine the parameter state simulation value as:

[0013] Among them, is the state simulation value of the output state of the i-th parameter at the k-th capture, y ik and P(y ik ) respectively correspond to the data state value taken by the output state of the i-th parameter at the k-th capture and the probability value of its occurrence; Step S24: Assign the data capture round number knum to knum + 1. If knum < TotalNum, enter Step S23; otherwise, enter Step S25; ​Step S25: Reconstruct the output state dataset of the digital virtual active suspension system to obtain a new analog output state dataset. , , This represents the simulated output state for the i-th output state. based on{ }Sure, ;Calculate the simulated output state The corresponding entropy matrix EN i .

[0014] Preferably, with Given the number of sub-segments m as the step size, calculate the entropy matrix EN. i ,in:

[0015] This indicates rounding down, λ is the dynamic capture coefficient, and T is the current time.

[0016] Preferably, the dynamic capture coefficient λ is determined based on the time range [0, T]. .

[0017] Preferably, step S3 includes: Step S31: Determine each entropy matrix EN i The extreme values ​​are used to determine the entropy matrix EN. i Each entropy element EN ij The corresponding positional probability P ij Based on location probability P ij Determine the weights for the distances between elements with entropy values; Among them, each entropy matrix EN i The extreme values ​​are:

[0018]

[0019] ; Step S32: Based on the weights of the distance between entropy elements, perform... Perform feature extraction and calculate the extracted feature and entropy matrix EN. i distance d i,j The entropy matrix EN is obtained. i The corresponding entropy element distance matrix; in:

[0020] Obtain the entropy matrix EN iThe corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T ; Step S33: Based on the entropy matrix EN i The corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T Determine a single multidimensional mixed indicator Ω.

[0021] In the formula, This is the correction coefficient. Distance value d i,j Relative to the entropy element distance matrix d The probability of occurrence; Step S34: Based on the objective function, perform feedback optimization on the values ​​of each parameter in the output state dataset of the digital virtual active suspension system; The objective function is:

[0022] in, The output of the vehicle parameters in the digital virtual active suspension system is mapped based on the vehicle state feedback, where ΔT is the system simulation prediction step size under digital mapping conditions, and e R Based on the location probability P ij The error correction value is calculated in real time using the virtual step size ΔT, where i=1, 2,…, 5; j=1, 2,…, 1000T-m+1; the virtual step size ΔT is calculated as follows: .

[0023] The present invention provides a device for dynamic design of active suspension parameters under digital mapping, the device comprising: System building module: Configured to set suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H roll Initial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θr and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Reconstruction module: Configured to obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Optimization module: configured to optimize each Perform feature extraction and calculate the extracted feature and entropy matrix EN. i The distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function.

[0024] The present invention provides a computer-readable storage medium storing a plurality of instructions; the plurality of instructions are used by a processor to load and execute the method as described above.

[0025] The present invention provides an electronic device, characterized in that the electronic device comprises: A processor is used to execute multiple instructions; Memory, used to store multiple instructions; The plurality of instructions are to be stored in the memory and loaded and executed by the processor as described above.

[0026] Beneficial effects: This invention constructs a virtual active suspension system and a digitized road environment under digital mapping, thereby dynamically adjusting and optimizing the parametric characteristics of the active suspension based on the vehicle's active suspension state output under preset initial parameter inputs. Specifically, a digitally simulated active suspension system is constructed through parameter selection and environmental data fusion, resulting in an output state dataset of the suspension system. Based on this, the suspension state output dataset is reconstructed by calculating the effective capture count n and the dynamic capture coefficient λ, thus obtaining a new simulated output state dataset. ; Calculate the dataset The corresponding entropy matrix, its element distance matrix d, and the correction coefficient κ can generate a single multi-dimensional hybrid index Ω. Finally, the performance target optimization function f can be set to optimize the feedback of the active suspension system parameters, thereby realizing the dynamic optimization and design of active suspension parameters in a digital virtual environment.

[0027] It has the following technical effects: (1) The present invention can overcome the problems of unpredictability, high cost, low reliability and unstable security in traditional solutions.

[0028] (2) The present invention obtains a digital simulation active suspension system in a virtual environment and performs multi-dimensional fusion of road environment information, thereby realizing system fusion and organic interaction of multiple information in a virtual environment.

[0029] (3) Based on the system fusion dataset in the virtual environment, the present invention resamples and randomly captures the vehicle state, thereby refitting the data distribution characteristics of the system and calculating the dynamic capture coefficient, thereby adjusting and calculating the dynamic range.

[0030] (4) This invention uses state entropy information to calculate the entropy position probability and the distance between sub-elements, thereby obtaining the reconstructed system output, and proposes a multi-dimensional information mixing index, thereby realizing active suspension parameter feedback optimization based on performance target optimization function. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the components of an active suspension system in the prior art; Figure 2 A schematic diagram of the dynamic design method for active suspension parameters under digital mapping provided by the present invention; Figure 3 A schematic diagram of the dynamic design architecture for active suspension parameters under digital mapping provided by the present invention; Figure 4 A schematic diagram of the structure of the active suspension parameter dynamic design device under digital mapping provided by the present invention. Detailed Implementation

[0032] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0033] like Figures 2-3 As shown, this invention proposes a dynamic design method for active suspension parameters under digital mapping, comprising the following steps: Step S1: Set the suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H rollInitial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Step S2: Obtain the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Step S3: For each Perform feature extraction and calculate the extracted feature and entropy matrix EN. i The distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function.

[0034] Step S1 includes: Step S11: Based on the current load of the vehicle and the requirements of the design objectives, set the suspension stiffness coefficient K, suspension damping coefficient C, electromagnetic thrust coefficient kF, suspension nonlinearity coefficient ξ, and vehicle roll center height H. roll The initial values ​​are used to form the parameter dataset x=[x1, x2, x3, x4, x5] of the digital virtual active suspension system. T =[K, C, k F , ξ, H roll ] T .

[0035] At this point, a digital virtual active suspension system at the current time T is obtained by using digital mapping.

[0036] Step S12: Calculate the road adhesion coefficient μ and the road surface roughness coefficient G. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φr As external environmental data, a road environment is constructed for the digital virtual active suspension system; the variation range of the road adhesion coefficient μ is set to (0, 1), the variation range of the vehicle speed V is set to [0, 150km / h], and the road roughness coefficient G... q The range of variation is set to [16, 16384].

[0037] In this embodiment, a virtual environment under multiple working conditions is digitally constructed. The road adhesion coefficient μ and road surface roughness coefficient G in the real environment are used. q The vehicle speed information V, vehicle travel path information S, and road slope angle θ measured by sensors in off-road environments are all included. r and roll angle φ r This data needs to be synchronously input into the system cloud platform as an external environment input dataset to complete the digital construction of the road environment in the virtual environment. Among them, the variation range of the road adhesion coefficient μ is set to (0, 1), the variation range of the vehicle speed V is set to [0, 150km / h], and the variation range of the road roughness coefficient Gq is set to [16, 16384].

[0038] Step S2 includes: Step S21: Based on the parameter dataset x of the digital virtual active suspension system and the road environment of the digital virtual active suspension system, determine the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T As state parameters representing the output state of the digital virtual active suspension system; obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T Construct the output state dataset y of the digital virtual active suspension system. y =[ y 1, y 2, y 3, y 4, y 5] T =[ D , , , v s , F T ] T Among them, under stable vehicle conditions, the output variable y i There is a corresponding safety threshold range H i ={[yimin , y imax ]}, output variable y i It is discrete, multidimensional data, with different dimensions representing system sampling data at different times. .

[0039] Furthermore, based on the data sampling frequency, the probability of the output state value corresponding to each output variable occurring can be determined, denoted as . P ( y iu ), u =1,2,…,1000 T , u These are variables related to the sampling frequency and time.

[0040] For example, if the sampling frequency is 1000Hz during data sampling, then at time T, the output state matrix y i The data dimension corresponding to (i=1,2,…,5) is 1000T, so the matrix dimension of the state dataset y is 5×1000T. Furthermore, based on the distribution of each parameter in the dataset, the probability of each output state value can be calculated as P(y). iu )(i=1,2,…,5; u=1,2,…,1000T).

[0041] Step S22: Set the number of data fetching rounds TotalNum Initialize the number of data fetching rounds knum to 0, and set... TotalNum It equals 1000T, where T is the current time; Step S23: For each output variable y i Based on the output variable y i There is a corresponding safety threshold range H i ={[y imin ,y imax ]} and output variable y i Determine the effective capture count n of the output state dataset of the digital virtual active suspension system, and n <1000 T ;

[0042] in, For y i H exceeds the safety threshold range i The amount of data, To round up; For each output variable y i The data in the sample is randomly sampled n times to determine the simulated values ​​of the parameter states. for:

[0043] Among them, is the state simulation value of the output state of the i-th parameter during the k-th capture, y ik and P(y ik ) respectively correspond to the data state value obtained by the output state of the i-th parameter during the k-th capture and the probability value of its occurrence.

[0044] For example, by randomly capturing data from the output states y i (i = 1, 2,..., 5), the number of data captures each time can be n (n < 1000T). The selection of the number of data captures n needs to be calculated according to the vehicle output state value y and its safety threshold range H described in step S21 under this working condition. After each data capture is completed, the captured data is put back into the original data set in the original order.

[0045] Assume that in the output variable matrix y i (i = 1, 2,..., 5), the number of states exceeding its safety threshold range H is n i (i = 1, 2,..., 5), and there is n i ≤1000T. Then, the corresponding effective number of captures n is rounded up, and we have:

[0046] In the formula, refers to the rounding-up operation performed on the decimal.

[0047] Thus, the parameter state simulation value under this data capture can be calculated as :

[0048] In the formula, y ik and P(y ik ) respectively correspond to the data state value obtained by the i-th output state during the k-th (k = 1, 2,..., n i ) capture and the probability value of its occurrence.

[0049] Step S24: Assign the data capture round number knum as knum + 1. If knum < TotalNum, enter step S23; otherwise, enter step S25; Step S25: Reconstruct the output state data set of the digital virtual active suspension system to obtain a new simulated output state data set , , is the simulated output state of the i-th output state, Based on {}Sure, ;Calculate the simulated output state The corresponding entropy matrix EN i .

[0050] Furthermore, with Given the number of sub-segments m as the step size, calculate the entropy matrix EN. i ,in:

[0051] This indicates rounding down, λ is the dynamic capture coefficient, and T is the current time.

[0052] Furthermore, the dynamic capture coefficient λ is determined based on the time range [0, T].

[0053] For example, by repeatedly capturing data 1000T times, a new simulated output state dataset for the digital suspension system can be obtained in the cloud platform. In the new data matrix In (i=1,2,…,5), calculate the corresponding entropy matrix EN according to a step size of m for the number of sub-segments. i (i=1,2,…,5), with a corresponding data dimension of 1×(1000T-m+1). The simulated output state dataset... The corresponding entropy matrix is ​​EN=[EN1, EN2, EN3, EN4, EN5]T, with a data dimension of 5×(1000T-m+1). The length m of the data sub-segment needs to be determined based on the output state matrix y. i The dynamic floor function is performed using the dimension (i=1,2,…,5), and the calculation method is as follows:

[0054] In the formula, The pointer performs a floor operation on the decimal ·, where λ is the dynamic fetching coefficient.

[0055] The selection of the dynamic capture coefficient λ needs to be dynamically calculated based on the extension of time T. Therefore, the coefficient λ is dynamically adjusted and calculated according to the time range [0, T], and the corresponding calculation formula is as follows:

[0056] Step S3 includes: Step S31: Determine each entropy matrix EN i The extreme values ​​are used to determine the entropy matrix EN. iEach entropy element EN ij The corresponding positional probability P ij Based on location probability P ij Determine the weights for the distances between elements with entropy values; Among them, each entropy matrix EN i The extreme values ​​are:

[0057]

[0058] .

[0059] For example, based on the entropy matrix EN i The interval weights are calculated using the set of (i=1, 2,…, 5). First, the five reconstructed output state variable datasets need to be calculated. Each corresponding entropy matrix EN i The extreme values ​​of (i=1,2,…,5) are shown below:

[0060] Based on the obtained entropy extreme value Max corresponding to each state variable. i and Min i Further solve the entropy matrix EN i Each entropy element EN in (i=1, 2,…, 5) ij The positional probability P corresponding to (i=1, 2,…, 5; j=1, 2,…, 1000T-m+1) ij That is, we have the following equation:

[0061] Based on the position probability P ij Based on the calculation results, a decision is made: if P ≥ 0.95, then the weight value is w1, and its range is set to [3, 4); if 0.75 ≤ P ≤ 0.95, then the weight value is w2, and its range is set to [2, 3); if 0 ≤ P ≤ 0.75, then the weight value is w3, and its range is set to [1, 2).

[0062] Step S32: Based on the weights of the distance between entropy elements, apply the weights to each... Perform feature extraction and calculate the extracted feature and entropy matrix EN. i distance d i,j The entropy matrix EN is obtained. i The corresponding entropy element distance matrix; in:

[0063]

[0064] Obtain the entropy matrix EN i The corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T .

[0065] For example, based on the element EN in the entropy matrix ij The positional probability P of (i=1,2,…,5; j=1,2,…,1000T-m+1) ij The corresponding weight value w k The value (k=1, 2, 3) can be used to adjust the system output parameters. Feature extraction is performed on the subsets (i=1, 2,…, 5). The entropy matrix EN is then used for calculation. i (i=1, 2,…, 5), calculate the distance d corresponding to each element. i,j The following is:

[0066] Therefore, the entropy matrix EN can be obtained. i The entropy element distance matrix corresponding to (i=1, 2,…, 5) is d =[ d 1, d 2, d 3, d 4, d 5] T The corresponding data dimension is 5×(1000T-m+1).

[0067] Step S33: Based on the entropy matrix EN i The corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T Determine a single multidimensional mixed indicator Ω.

[0068] In the formula, This is the correction coefficient. Distance value d i,j Relative to the entropy element distance matrix dThe probability of occurrence.

[0069] Furthermore, the correction coefficient The selection and calculation of the value need to consider the changing trend of the vehicle speed V; the specific method for obtaining the correction coefficient κ value is as follows:

[0070] In the formula, R is the effective radius of the tire during vehicle travel, and ω is the angular velocity of the wheel.

[0071] Step S34: Based on the objective function, perform feedback optimization on the values ​​of each parameter in the output state dataset of the digital virtual active suspension system; The objective function is:

[0072] in, The output of the vehicle parameters in the digital virtual active suspension system is mapped based on the vehicle state feedback, where ΔT is the system simulation prediction step size under digital mapping conditions, and e R Based on the location probability P ij The error correction value is calculated in real time using the virtual step size ΔT, where i=1, 2,…, 5; j=1, 2,…, 1000T-m+1; the virtual step size ΔT is calculated as follows:

[0073] In this embodiment, in the digital virtual active suspension system, based on the state output matrix and error feedback correction of the digital simulation active suspension system, the multi-dimensional hybrid index Ω can be minimized through parameter calculation and optimization, thereby realizing the dynamic optimization design of active suspension parameters under digital mapping conditions.

[0074] This invention also provides a device for dynamic design of active suspension parameters under digital mapping, such as... Figure 4 As shown, the device includes: System building module: Configured to set suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H roll Initial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Reconstruction module: Configured to obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Optimization module: configured to optimize each Perform feature extraction and calculate the extracted feature and entropy matrix EN. i The distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function.

[0075] The specific embodiments described above only illustrate the design principles of the present invention. The shapes and names of the components in this description may differ and are not limited. Therefore, those skilled in the art can modify or make equivalent substitutions to the technical solutions described in the foregoing embodiments; and these modifications and substitutions do not depart from the inventive spirit and technical solutions of the present invention, and should all fall within the protection scope of the present invention.

Claims

1. A method for dynamic design of active suspension parameters under digital mapping, characterized in that, It includes the following steps: Step S1: Set the suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H roll Initial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Step S2: Obtain the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Step S3: For each Perform feature extraction and calculate the extracted feature and entropy matrix EN. i The distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function; The step S2 includes: Step S21: Based on the parameter dataset x of the digital virtual active suspension system and the road environment of the digital virtual active suspension system, determine the suspension dynamic deflection D and the sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T As state parameters representing the output state of the digital virtual active suspension system; obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T Construct the output state dataset y of the digital virtual active suspension system. y =[ y 1, y 2, y 3, y 4, y 5] T =[ D , , , v s , F T ] T Among them, under stable vehicle conditions, the output variable y i There is a corresponding safety threshold range H i ={[y imin , y imax ]}, output variable y i It is discrete, multidimensional data, with different dimensions representing system sampling data at different times. ; Step S22: Set the number of data fetching rounds TotalNum Initialize the number of data fetching rounds knum to 0, and set... TotalNum It equals 1000T, where T is the current time; Step S23: For each output variable y i Based on the output variable y i There is a corresponding safety threshold range H i ={[y imin , y imax ]} and output variable y i Determine the effective capture count n of the output state dataset of the digital virtual active suspension system, and n <1000 T ; in, For y i H exceeds the safety threshold range i The amount of data, To round up; For each output variable y i The data in the sample is randomly sampled n times to determine the simulated values ​​of the parameter states. for: in, Let y be the simulated state value of the output state of the i-th parameter during the k-th round of capture. ik and P(y ik The output states of the i-th parameter correspond to the data state value and its probability value obtained in the k-th round of crawling, respectively. Step S24: Assign the data scraping round number knum as knum plus 1. If knum < TotalNum, enter step S23; otherwise, enter step S25; Step S25: Reconstruct the output state dataset of the digital virtual active suspension system to obtain a new simulated output state dataset. , , This represents the simulated output state for the i-th output state. based on{ }Sure, ;Calculate the simulated output state The corresponding entropy matrix EN i .

2. The method as described in claim 1, characterized in that, The step S1 includes: Step S11: Based on the current load of the vehicle and the requirements of the design objectives, set the suspension stiffness coefficient K, suspension damping coefficient C, electromagnetic thrust coefficient kF, suspension nonlinearity coefficient ξ, and vehicle roll center height H. roll The initial values ​​are used to form the parameter dataset x=[x1, x2, x3, x4, x5] of the digital virtual active suspension system. T =[K, C, k F , ξ, H roll ] T ; Step S12: Calculate the road adhesion coefficient μ and the road surface roughness coefficient G. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r As external environmental data, a road environment is constructed for the digital virtual active suspension system; the variation range of the road adhesion coefficient μ is set to (0, 1), the variation range of the vehicle speed V is set to [0, 150km / h], and the road roughness coefficient G... q The range of variation is set to [16, 16384].

3. The method as described in claim 2, characterized in that, by Given the number of sub-segments m as the step size, calculate the entropy matrix EN. i ,in: This indicates rounding down, λ is the dynamic capture coefficient, and T is the current time.

4. The method as described in claim 3, characterized in that, The dynamic scraping coefficient λ is determined according to the size of the time range [0, T], 。 5. The method according to any one of claims 2-4, characterized in that, The step S3 includes: Step S31: Determine each entropy matrix EN i The extreme values ​​are used to determine the entropy matrix EN. i Each entropy element EN ij The corresponding position probability P ij Based on location probability P ij Determine the weights for the distances between elements with entropy values; Among them, each entropy matrix EN i The extreme values ​​are: ； Step S32: Based on the weights of the distance between entropy elements, apply the weights to each... Perform feature extraction and calculate the extracted feature and entropy matrix EN. i distance d i,j The entropy matrix EN is obtained. i The corresponding entropy element distance matrix; Where: The entropy matrix EN is obtained. i The corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T , The weights for the distance between elements of entropy value; Step S33: Based on the entropy matrix EN i The corresponding entropy element distance matrix d =[ d 1, d 2, d 3, d 4, d 5] T Determine a single multidimensional mixed indicator Ω. In the formula, This is the correction coefficient. Distance value d i,j Distance matrix relative to entropy elements d The probability of occurrence; Step S34: Based on the objective function, perform feedback optimization on the numerical values of each parameter in the output state data set of the digital virtual active suspension system; The objective function is: in, The output of the vehicle parameters in the digital virtual active suspension system is mapped based on the vehicle state feedback, where ΔT is the system simulation prediction step size under digital mapping conditions, and e R Based on the location probability P ij The error correction value is calculated in real time using the virtual step size ΔT, where i=1, 2,…, 5; j=1, 2,…, 1000T-m+1; the virtual step size ΔT is calculated as follows: 。 6. A device for dynamic design of active suspension parameters under digital mapping, used to execute the method of any one of claims 1-5, characterized in that, The device includes: System building module: Configured to set suspension stiffness coefficient K, suspension damping coefficient C, and electromagnetic thrust coefficient k. F Suspension nonlinearity coefficient ξ and vehicle roll center height H roll Initial values ​​are used to construct the parameter dataset x of the digital virtual active suspension system; the road adhesion coefficient μ and the road surface roughness coefficient G are used to construct the parameter dataset x of the digital virtual active suspension system. q Vehicle speed information V, driving route information S, and road slope angle θ r and roll angle φ r Using external environmental data, we construct the road environment for a digital virtual active suspension system. Reconstruction module: Configured to obtain suspension dynamic deflection D and sprung mass acceleration. Unsprung mass acceleration Suspension speed v s and tire dynamic load F T The output state dataset y of the digital virtual active suspension system and the safety threshold range of each parameter in dataset y are constructed. The number of wheels to be captured in the output state dataset of the digital virtual active suspension system is determined. Then, the output state dataset of the digital virtual active suspension system is reconstructed to obtain a new simulated output state dataset. ,calculate Each element The corresponding entropy matrix EN i , ; Optimization module: configured to optimize each Perform feature extraction and calculate the extracted feature and entropy matrix EN. i The distance; determine a single multi-dimensional hybrid index, establish an objective function based on the multi-dimensional hybrid index, and optimize the values ​​of each parameter in the output state dataset of the digital virtual active suspension system based on the objective function.

7. A computer-readable storage medium, in which multiple instructions are stored; the multiple instructions are used to be loaded and executed by a processor to perform the method according to any one of claims 1-5.

8. An electronic device, characterized in that, The electronic device includes: A processor for executing multiple instructions; A memory for storing multiple instructions; Among them, the multiple instructions are used to be stored by the memory and loaded and executed by the processor to perform the method according to any one of claims 1-5.

Images