Real-time prediction method and controller for optimal slip ratio of vehicle based on ratio division

Abstract

The present application belongs to the field of vehicle control, and particularly relates to a real-time prediction method and controller for optimal slip rate of a vehicle based on ratio division. The method first constructs an attached slip curve of the vehicle under different road conditions, and determines the corresponding optimal slip rate, then takes the ratio of the real-time slip rate in the attached slip curve and the optimal slip rate as the slip rate ratio, and fits a plurality of ratio curves of gradient values corresponding to different ratios; then real-time collects the wheel speed information of the current vehicle, and combines the dynamics model of the vehicle to generate the utilization of the attached coefficient and the slip rate of the vehicle in real time, forming the current vehicle state; finally, the ratio curve closest to the vehicle state is screened out, and then the ratio of the real-time slip rate and the gradient value of the ratio curve is taken as the optimal slip rate at the current time. Compared with the existing scheme, the present application has higher prediction accuracy and efficiency, and better real-time performance; and can solve the problems of insufficient prediction accuracy and high delay of the optimal slip rate in the existing longitudinal stability control of electric vehicles.

Description

Technical Field

[0001] This invention belongs to the field of vehicle control, specifically relating to a real-time prediction method for the optimal vehicle slip ratio based on ratio division, and the corresponding vehicle longitudinal stability control method, vehicle longitudinal stability controller, and vehicle. Background Technology

[0002] With the continuous advancement of electronic and intelligent technologies, electric vehicles are becoming increasingly mature, and their market share is gradually expanding. Compared to gasoline vehicles, electric vehicles have enormous potential in reducing energy consumption, improving driving safety, and enhancing driving performance. Because electric vehicles use an electric motor-based braking system, rather than a hydraulic or pneumatic braking system, they offer superior real-time performance and rapid response, eliminating hydraulic or mechanical delays that could hinder precise control. Therefore, research on the longitudinal stabilization system of electric vehicles is of great significance.

[0003] The design of the braking longitudinal stability system for electric vehicles mainly involves road type identification and prediction, which can be divided into cause-based methods and response-based methods. Cause-based methods require the measurement of external information such as optical, acoustic, temperature, and humidity. This method offers more accurate predictions but is costly, has poor robustness, and struggles to accurately estimate untested roads. Response-based methods primarily calculate the maximum road adhesion coefficient using a dynamic model during vehicle braking. This approach offers more accurate identification but is difficult to control in real-time and exhibits a certain degree of lag.

[0004] The design of a longitudinal braking stability system essentially involves controlling the vehicle to maximize the utilization of the coefficient of friction, thereby achieving maximum longitudinal braking force under conditions of maximum road surface friction. Achieving maximum road surface friction equates to achieving the optimal slip ratio. By controlling the slip ratio during braking to reach this optimal slip ratio, efficient braking performance can be obtained. Causal methods require excessive measurement data and processing time, making it impossible to obtain the optimal slip ratio at the current moment or a specific point in time, thus lacking real-time capability. Response-based methods, such as maximum road surface friction identification, require the vehicle to have reached or be about to reach the critical state of braking instability, often resulting in a sampling time lag. Therefore, predicting the optimal slip ratio before the vehicle reaches the critical state of braking instability, and thus enabling precise vehicle control, has become a pressing technical challenge for those skilled in the art. Summary of the Invention

[0005] To address the issues of insufficient accuracy and significant delay in predicting the optimal slip ratio in existing electric vehicle longitudinal stability control systems, this invention provides a real-time prediction method for the optimal slip ratio of a vehicle based on ratio partitioning, along with a corresponding vehicle longitudinal stability control method, a vehicle longitudinal stability controller, and a vehicle.

[0006] This invention is achieved using the following technical solution:

[0007] A real-time prediction method for optimal vehicle slip ratio based on ratio partitioning, comprising:

[0008] Simultaneously collect vehicle adhesion coefficient under various road conditions With slip ratio Use it as road test data ( , Based on a large amount of road test data collected for each road condition, corresponding slip curves are fitted, and the optimal slip ratio for each slip curve is determined. This allows us to obtain multiple slip curves and their optimal slip ratios for the current vehicle under different road conditions.

[0009] The ratio of the real-time slip ratio to the optimal slip ratio in the slip curve is used as the slip ratio ratio, thereby determining the slip ratio ratio for each data point in the slip curve. The slip ratio ratio is then divided into multiple gradient values ​​according to a preset accuracy level. Points with the same slip ratio on all slip curves are fitted to obtain the ratio curve corresponding to the gradient values.

[0010] Real-time acquisition of vehicle wheel speed information, combined with the vehicle's dynamics model, generates the vehicle's adhesion coefficient in real time. and slip ratio This constitutes the current vehicle status. , From the various ratio curves, select the ratio curve that best matches the vehicle state, and then calculate the real-time slip ratio. The ratio of the gradient value of the ratio curve is taken as the optimal slip ratio at the current moment.

[0011] As a further improvement of the present invention, the vehicle's real-time adhesion coefficient The calculation formula is as follows:

[0012] ;

[0013] In the above formula, F zr This is the vertical force on the rear wheel; F xr This is the longitudinal force of the rear wheel; L This refers to the wheelbase between the front and rear wheels; G For vehicle weight; a This is the longitudinal distance from the vehicle's center of gravity to the front wheel; m For vehicle quality; h g The height of the vehicle's center of gravity; To accelerate the vehicle; This refers to the angular acceleration of the front wheel; rThe radius of the wheel; I Let be the moment of inertia of the wheel.

[0014] As a further improvement of the present invention, the formula for calculating the real-time slip rate of the vehicle is as follows:

[0015] ;

[0016] In the above formula, v For vehicle speed; This indicates the angular velocity of the rear wheel.

[0017] As a further improvement of this invention, multiple slip curves and their corresponding ratio curves under different road conditions are plotted on the same image as a working condition atlas; in the vehicle optimal slip rate update task, based on the vehicle state ( , ) Query the operating condition graph, determine the closest ratio curve, and calculate the vehicle's current optimal slip ratio.

[0018] As a further improvement of the present invention, the working condition map is dynamically updated. When the vehicle is driving under new road conditions, a new slip curve is fitted based on the real-time updated vehicle state value, and each ratio curve is dynamically updated based on the new slip curve.

[0019] As a further improvement of the present invention, the ratio curve closest to the current vehicle state is determined by calculating the minimum Euclidean distance between the current vehicle state value and each ratio curve in the working condition graph.

[0020] Alternatively, the slip curve and ratio curve in the working condition diagram can be represented in matrix form, and the ratio curve that is closest to the current vehicle state can be obtained through an arithmetic optimization algorithm.

[0021] As a further improvement of the present invention, the method for solving the ratio curve that is closest to the current vehicle state based on the arithmetic optimization algorithm includes:

[0022] i. By fitting the slip curve with a polynomial function, then in the th... i On the slip curve, the ratio is p point ( , The following is represented:

[0023] ;

[0024] In the above formula, a i0 ~ a im The constant terms in the slip curve are respectively... m The coefficient of the second term; For the first iThe optimal slip ratio on the slip curve; p This represents the ratio of real-time slip rate to the optimal slip rate. For the first i The ratio on the slip curve is p The point corresponds to the adhesion coefficient.

[0025] ii. Let the number of attached sliding curves be n Then each slip curve has the same ratio p Each point ( , ), ( , ), ..., ( , The following is represented:

[0026] ;

[0027] In the above formula, ; .

[0028] iii. For different slip curves with the same ratio p By fitting the data to each point, the corresponding ratio curve is obtained as follows:

[0029] ;

[0030] In the above formula, s 0~ s n-1 These are the constant terms in the ratio curve to... n The coefficient of the -1 term.

[0031] iv. Different slip curves have the same ratio p The points are represented as follows:

[0032] .

[0033] v. When any ratio curve is closest to the current vehicle state, the current vehicle's real-time utilization of the adhesion coefficient is... Corresponding to real-time slip rate The cost function is constructed by minimizing the difference in the adhesion coefficients of the ratio curves. D 1 is:

[0034] .

[0035] vi. Preset ratio p The search range is determined by the ratio using an arithmetic optimization algorithm. p The cost function is optimized to minimize it, and this value is used as the ratio value of the ratio curve that is closest to the current vehicle state.

[0036] As a further improvement to this invention, an arithmetic optimization algorithm is used to compare the ratio. p The iterative optimization process includes:

[0037] I. Initialize a uniformly distributed ratio using the following formula. p The initial population:

[0038] ;

[0039] In the above formula, Ub For ratio p The upper bound of the range of values; Lb For ratio p Lower bound of the value range; Rand Represents a random number between 0 and 1; The first in the population i The individual in the first j The location in 3D space.

[0040] II. Use the Mathematical Function Accelerator (MOA) to determine whether to perform global exploration or local development. The MOA expression is:

[0041] ;

[0042] In the above formula, MOA( t () represents the current iteration number. t The acceleration function value; Min This represents the minimum value of the acceleration function; Max This represents the maximum value of the acceleration function; T This represents the maximum number of iterations.

[0043] III. The probability of exploration and development is controlled using the mathematical optimizer (MOP). The calculation formula is as follows:

[0044] ;

[0045] In the above formula, MOP( t ) represents the current mathematical optimizer probability; For iteration sensitivity coefficient, A higher value results in a higher number of iterations for MOP ( t The greater the impact, the better.

[0046] IV. When random numbers r If 1 < MOA, proceed to the exploration phase; otherwise, proceed to the development phase.

[0047] During the exploration phase, based on random numbers r 2. Decide whether to use a multiplication or division strategy to update the population. The update formula is:

[0048] ;

[0049] In the above formula, r 2 is a random number between 0 and 1; X( t +1) represents the updated position of the individual; X b ( t () represents the position of the individual with the best fitness before the update; For the control coefficient of the search process; This is to adjust the minimum value of the probability.

[0050] During the development phase, based on random numbers r 3. Decide whether to use an addition or subtraction strategy to update the population. The update formula is:

[0051] .

[0052] V. Iteratively update the population and calculate the cost function value of each individual in the population in each round as its fitness value. Select the individual with the smallest fitness value as the optimal individual in each round. When the preset maximum number of iterations is reached, output the optimal individual as the optimization result.

[0053] The present invention also includes a vehicle longitudinal stability control method based on an optimal slip ratio, which includes the following steps:

[0054] S1: The optimal slip ratio of the vehicle is generated in real time using the aforementioned ratio-based real-time prediction method. .

[0055] S2: Using the optimal slip ratio from the previous step as the target slip ratio, construct the state-space equations for the vehicle longitudinal stability control strategy as follows:

[0056] ;

[0057] In the above formula, The first derivative representing the increment of slip ratio; This represents the increment of the slip ratio; This represents the torque increment.

[0058] S3: Will As vehicle status x Then the state-space equations are transformed into standard form:

[0059] ;

[0060] In the above formula, Indicates vehicle state increment The first derivative; This indicates the control increment. ;A c The coefficients of the state and control terms are represented by . ; .

[0061] After discretization, it is represented as follows:

[0062] ;

[0063] Among them, △ x [ k ] and △ x [ k+ 1] respectively represent k and k Vehicle status at time +1; △ u [ k ]express k The torque increment at any given moment.

[0064] S4: Setting the cost function corresponding to the vehicle longitudinal stability control strategy based on the discretized state-space equations. J for:

[0065] ;

[0066] In the above formula, q 1 and q 2 represents the weighting coefficients for the status term and the control term, respectively; y [ k + j | k ]express k Time prediction k + j Vehicle status at any given time; y r [ k + j ]express k + j The target vehicle status at any given time; △ u [ k + j ]express k + j The torque increment at any given moment; p t Indicates the state step size; m t This indicates the control step size.

[0067] S5: Using any iterative optimization algorithm, based on the current state of the vehicle, iteratively optimize the cost function to achieve the desired result. J Minimize the torque increment at the next moment and use it to regulate the motor output torque of the vehicle.

[0068] The present invention also includes a vehicle longitudinal stability controller, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the vehicle longitudinal stability control method based on the aforementioned optimal slip ratio, and then performs longitudinal stability control on the vehicle in real time based on the collected vehicle wheel speed information and output torque.

[0069] The present invention also includes a vehicle that employs the vehicle longitudinal stability controller as described above.

[0070] The technical solution provided by this invention has the following beneficial effects:

[0071] This invention is the first to discover the similarity of geometric features between points with the same ratio in different slip curves, and proposes a real-time prediction method for the optimal vehicle slip ratio based on ratio division. This method only needs to obtain the vehicle's real-time utilization coefficient of adhesion and slip ratio based on the vehicle's wheel speed information to query a pre-constructed working condition map and obtain the optimal slip ratio under the current vehicle state. Compared with existing solutions, the new solution of this invention has higher prediction efficiency and better real-time performance while maintaining prediction accuracy. Furthermore, the working condition map can be dynamically updated to further improve prediction accuracy.

[0072] This invention also utilizes an extended state observer combined with model predictive control for braking control. Using the aforementioned optimal slip ratio as a reference, a model predictive algorithm is employed to solve for the control variable that minimizes the cost function. This variable is then input into the motor for control, adjusting the braking system. This achieves the goal of reducing braking distance and ensuring passenger safety. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the working condition map constructed in Embodiment 1 of the present invention.

[0074] Figure 2 This is a flowchart illustrating the steps of a real-time prediction method for the optimal vehicle slip ratio based on ratio division in Embodiment 1 of the present invention.

[0075] Figure 3 This is the dynamic model for vehicle longitudinal stability control constructed in Embodiment 1 of the present invention.

[0076] Figure 4 This is a flowchart of the iterative optimization of the ratio using an arithmetic optimization algorithm in Embodiment 1 of the present invention.

[0077] Figure 5 This is a schematic diagram of a vehicle longitudinal stability controller including an extended state observer provided in Embodiment 3 of the present invention.

[0078] Figure 6The graph shows the changes in the adhesion coefficient, speed, slip ratio, and motor output torque of the vehicle using the solution of this invention over time during a simulation experiment. Detailed Implementation

[0079] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0080] Example 1

[0081] Slip ratio is a parameter characterizing the proportion of slippage during vehicle movement, including slip ratio under driving conditions and slip ratio under braking conditions. When any vehicle travels on different road surfaces, the coefficient of adhesion varies with the slip ratio due to differences in vehicle parameters and road condition parameters such as the coefficient of friction. In practical applications, corresponding slip curves can be plotted to evaluate the mapping relationship between the coefficient of adhesion and slip ratio for the current vehicle on different road surfaces. Furthermore, predicting the optimal slip ratio based on the real-time updated slip curve under the current driving state can provide a basis for controlling the vehicle's braking or driving operations under different driving states, thereby ensuring the vehicle's longitudinal stability.

[0082] In this embodiment, after conducting massive data analysis on the vehicle's slip curves and corresponding optimal slip ratios under different road conditions, the technicians found that: Figure 1 As shown, assuming the real-time slip ratio in any slip curve is... With optimal slip ratio The ratio is denoted as the slip ratio. p best (Hereinafter referred to as slip ratio):

[0083] ;

[0084] Next, points with the same ratio in different slip curves are fitted to obtain the corresponding ratio curves. Finally, observing the graphical characteristics of the ratio curves under different ratios reveals that each ratio curve exhibits very similar geometric features. Therefore, this ratio curve can be used to reflect the correlation between the stored curves. Under this condition, when a ratio curve passes through or approaches the real-time point of the current vehicle state ( , When ), it can be assumed that the point has the same ratio as the ratio curve on its own slip curve. Wherein, and These represent the vehicle's real-time slip rate and adhesion coefficient, respectively.

[0085] In other words, if a slip curve that is updated in real time according to the vehicle's status passes through or is close to a certain ratio curve, then it can be determined that the ratio between the real-time slip rate and the optimal slip rate of the vehicle in its current state should be approximately equal to the ratio corresponding to that ratio curve.

[0086] Based on the above findings, this embodiment provides a novel real-time prediction method for the optimal vehicle slip ratio based on ratio partitioning, such as... Figure 2 As shown, it includes the following steps:

[0087] 1. Simultaneously collect vehicle adhesion coefficient data under various road conditions. With slip ratio Use it as road test data ( , Based on a large amount of road test data collected for each road condition, corresponding slip curves are fitted, and the optimal slip ratio for each slip curve is determined. This allows us to obtain multiple slip curves and their optimal slip ratios for the current vehicle under different road conditions.

[0088] In practical applications, this embodiment can utilize a large amount of road test data collected during the design and testing phases of each vehicle to plot the vehicle's slip curves under different road conditions. Specifically, to ensure that the plotted slip curves comprehensively cover the vehicle's performance under all road conditions, each slip curve should include data from dry, rough roads where the road surface adhesion coefficient is highest, as well as data from wet, icy, or snowy road conditions where the road surface adhesion coefficient is lowest.

[0089] Second, the ratio of the real-time slip ratio to the optimal slip ratio in the slip curve is used as the slip ratio ratio, thereby determining the slip ratio ratio for each data point in the slip curve. The slip ratio ratio is divided into multiple gradient values ​​according to a preset accuracy level; points with the same slip ratio on all slip curves are fitted to obtain the ratio curve corresponding to the gradient values.

[0090] In practical applications, if multiple slip curves and their corresponding ratio curves under different road conditions are plotted on the same image, this image can be called a working condition map. Figure 1 The displayed diagram shows a typical operating condition for a specific vehicle. Clearly, the various slip curves and ratio curves in the operating condition diagram represent the entire slip plane (within...). for x Axis, with for y The plane formed by the axes is divided into multiple grids.

[0091] Third, the vehicle's wheel speed information is collected in real time, and the vehicle's adhesion coefficient is generated in real time based on the vehicle's dynamic model. and slip ratio This constitutes the current vehicle status. , ).

[0092] In order to obtain the vehicle's coefficient of adhesion based on the vehicle's wheel speed information and slip ratio The following describes the construction process of the vehicle's dynamic model in the longitudinal stability control of this embodiment:

[0093] Considering that factors such as lateral dynamics have a relatively small impact on this method, the vehicle model can be subjected to, for example... Figure 3 The simplification shown is as follows. For the vehicle dynamics model used for longitudinal stability control, the following assumptions can be made: (1) Since rolling resistance, air resistance, and road slope are not considered, this study only focuses on the longitudinal dynamics model of the vehicle. (2) The driving torque of the tires is provided by the motor in real time, and the reducer and other lag factors can be ignored. (3) The front wheels are not directly affected by the rear wheel drive motor. Therefore, the linear velocity of the front wheels as driven wheels can be approximated as the vehicle speed.

[0094] In such Figure 3 In the longitudinal dynamics model of the vehicle shown, the rear wheels of the vehicle satisfy the following torque balance formula:

[0095] ;

[0096] In the above formula, T r This refers to the torque on the rear wheels; F xr This is the longitudinal force of the rear wheel; This refers to the angular acceleration of the rear wheel; r The radius of the wheel; I Let be the moment of inertia of the wheel.

[0097] This study focuses on longitudinal stability control, including traction control and anti-lock braking. Under driving conditions, the longitudinal force on the rear wheels can be obtained from the longitudinal dynamics model. F xr and vertical force F zr Represented as:

[0098] ;

[0099] In the above formula, L This refers to the wheelbase between the front and rear wheels; G For vehicle weight; a This is the longitudinal distance from the vehicle's center of gravity to the front wheel; m For vehicle quality; h gThe height of the vehicle's center of gravity; To accelerate the vehicle; This refers to the angular acceleration of the front wheel; r The radius of the wheel; I Let be the moment of inertia of the wheel.

[0100] At this point, the coefficient of friction of the vehicle's rear wheels... Satisfy the following formula: .

[0101] Based on the previous equation, the torque on the rear wheel can be obtained. T r for: .

[0102] Under vehicle driving conditions, slip ratio The sign is set to negative, which is represented as: ;in, For vehicle speed; This is the angular velocity of the rear wheel.

[0103] The above formula can be transformed into: Differentiating both sides of the equation, we get:

[0104] ;

[0105] In the above formula, , and They are respectively , and v The first derivative.

[0106] Substituting into the previous equation, we get: Further analysis reveals: .

[0107] Combined with the torque on the rear wheels T r The equation yields:

[0108] .

[0109] Under braking conditions, slip ratio Setting it to positive means: This formula can be transformed into: .

[0110] Differentiating both sides of the equation, we get: .

[0111] Combining the previous equation, we can obtain: .

[0112] In summary, the vehicle's real-time coefficient of adhesion The calculation formula is as follows:

[0113] .

[0114] Real-time vehicle slip rate The calculation formula is as follows:

[0115] .

[0116] In this embodiment, the vehicle status is continuously collected at different times, thereby enabling... Figure 1 A series of data points are obtained on the slip plane shown. By connecting these data points sequentially along the time axis and fitting them to a curve, the slip curve of the vehicle under the current road conditions can be obtained. The end of this slip curve (i.e., the point corresponding to the current vehicle state) may intersect with a certain ratio curve, or lie between two ratio curves with different ratio values. Obviously, the ratio of the ratio curve that is closest to or intersects with the point corresponding to the current vehicle state is exactly close to the ratio of the real-time slip rate to the optimal slip rate of the current slip curve. If the density of the slip curve meets the requirements (i.e., the gradient value of the ratio is small enough), the two ratios can be considered equal.

[0117] Fourth, select the ratio curve that best matches the vehicle's condition from all the ratio curves, and then calculate the real-time slip ratio. The ratio of the gradient value of the ratio curve is taken as the optimal slip ratio at the current moment.

[0118] Assuming a real-time updated vehicle status ( , The ratio of the closest ratio curve is p best The optimal sliding rate at the current moment is :

[0119] .

[0120] In the practical application of this embodiment, the optimal slip ratio can be predicted using a lookup table method or a function method. Specifically, in the lookup table method, multiple slip curves and their corresponding ratio curves under different road conditions can be plotted on the same image as a working condition atlas. In the task of updating the vehicle's optimal slip ratio, based on the real-time updated vehicle state ( , The system queries the operating condition map to determine the closest ratio curve and calculates the vehicle's current optimal slip ratio. For example, based on the known functional equations of the fitted slip curves and ratio curves, the ratio curve closest to the current vehicle state can be determined by calculating the minimum Euclidean distance between the current vehicle state value and each ratio curve in the operating condition map.

[0121] Furthermore, in practical applications, when vehicles are traveling under entirely new road conditions, vehicle status data can be further collected. Based on the real-time updated vehicle status values, a new slip curve can be fitted. Then, based on the new slip curve, each ratio curve is dynamically updated, resulting in a more refined operating condition map. These updated curves make the lines in the operating condition map denser, the grid finer, and further improve the accuracy of the closest ratio curve found in the final query.

[0122] In the function method, this embodiment represents the slip curve and ratio curve in the working condition diagram as a matrix, and uses an arithmetic optimization algorithm to solve for the ratio curve that is closest to the current vehicle state. Specifically, the method for solving for the ratio curve that is closest to the current vehicle state based on the arithmetic optimization algorithm includes:

[0123] i. By fitting the slip curve with a polynomial function, then in the th... i On the slip curve, the ratio is p point ( , The following is represented:

[0124] ;

[0125] In the above formula, a i0 ~ a im The constant terms in the slip curve are respectively... m The coefficient of the second term; For the first i The optimal slip ratio on the slip curve; p This represents the ratio of real-time slip rate to the optimal slip rate. For the first i The ratio on the slip curve is p The point corresponds to the adhesion coefficient.

[0126] ii. Let the number of attached sliding curves be n Then each slip curve has the same ratio p Each point ( , ), ( , ), ..., ( , The following is represented:

[0127] ;

[0128] In the above formula, ; .

[0129] iii. For different slip curves with the same ratio p By fitting the data to each point, the corresponding ratio curve is obtained as follows:

[0130] ;

[0131] In the above formula, s 0~ s n-1 These are the constant terms in the ratio curve to... n The coefficient of the -1 term.

[0132] iv. Different slip curves have the same ratio p The points are represented as follows:

[0133] .

[0134] v. When any ratio curve is closest to the current vehicle state, the current vehicle's real-time utilization of the adhesion coefficient is... Corresponding to real-time slip rate The cost function is constructed by minimizing the difference in the adhesion coefficients of the ratio curves. D 1 is:

[0135] .

[0136] vi. Preset ratio p The search range is determined by the ratio using an arithmetic optimization algorithm. p The cost function is optimized to minimize it, and this value is used as the ratio value of the ratio curve that is closest to the current vehicle state.

[0137] Among them, such as Figure 4 As shown, the ratio is compared using an arithmetic optimization algorithm. p The iterative optimization process includes:

[0138] I. Initialize a uniformly distributed ratio using the following formula. p The initial population:

[0139] ;

[0140] In the above formula, Ub For ratio p The upper bound of the range of values; Lb For ratio p Lower bound of the value range; Rand Represents a random number between 0 and 1; The first in the population i The individual in the first j The location in 3D space.

[0141] II. Use the Mathematical Function Accelerator (MOA) to determine whether to perform global exploration or local development. The MOA expression is:

[0142] ;

[0143] In the above formula, MOA( t () represents the current iteration number. t The acceleration function value; Min This represents the minimum value of the acceleration function; Max This represents the maximum value of the acceleration function; T This represents the maximum number of iterations.

[0144] III. The probability of exploration and development is controlled using the mathematical optimizer (MOP). The calculation formula is as follows:

[0145] ;

[0146] In the above formula, MOP( t ) represents the current mathematical optimizer probability; For iteration sensitivity coefficient, A higher value results in a higher number of iterations for MOP ( t The greater the impact, the better.

[0147] IV. When random numbers r If 1 < MOA, proceed to the exploration phase; otherwise, proceed to the development phase.

[0148] During the exploration phase, based on random numbers r 2. Decide whether to use a multiplication or division strategy to update the population. The update formula is:

[0149] ;

[0150] In the above formula, r 2 is a random number between 0 and 1; X( t +1) represents the updated position of the individual; X b ( t () represents the position of the individual with the best fitness before the update; For the control coefficient of the search process; This is to adjust the minimum value of the probability.

[0151] During the development phase, based on random numbers r 3. Decide whether to use an addition or subtraction strategy to update the population. The update formula is:

[0152] .

[0153] V. Iteratively update the population and calculate the cost function value of each individual in the population in each round as its fitness value. Select the individual with the smallest fitness value as the optimal individual in each round. When the preset maximum number of iterations is reached, output the optimal individual as the optimization result.

[0154] The difference between the lookup table method and the function method lies in the following: In the lookup table method, the functional equations of each slip function and ratio curve are established based on existing road condition data. When the road condition data is updated, the relevant functional equations need to be refitted and updated. In the function method, both the slip function and ratio curve are fitted using polynomials and represented by matrices. When the road condition data is updated, only the dimensions of the correlation coefficient matrix and variable matrix need to be increased; there is no need to refit the corresponding equations. Therefore, the latter is more user-friendly for vehicle control systems in practical applications. For example, when designing the corresponding coefficient matrix and traversal matrix, engineers can separate the stored information from the real-time updated data, making it easier to directly incorporate this information into the solution after the data is updated.

[0155] Example 2

[0156] Based on the actual example scheme, this embodiment further provides a vehicle longitudinal stability control method based on the optimal slip ratio, which includes the following steps:

[0157] S1: The optimal slip ratio of the vehicle is generated in real time using the ratio-based vehicle optimal slip ratio prediction method as described in Example 1. .

[0158] S2: Use the optimal slip ratio from the previous step as the target slip ratio. In practical applications, longitudinal stability control is divided into driving and braking, with different control strategies corresponding to changes in driving intent. In this embodiment, the target slip ratio can be used as a condition for distinguishing intent; a positive sign indicates braking, and a negative sign indicates driving.

[0159] Under the driving condition, the incremental state-space equation is as follows:

[0160] ;

[0161] In the above formula: To drive the increment of slip ratio The first derivative; This corresponds to the increase in motor drive torque.

[0162] Under braking conditions, the incremental state-space equations are as follows:

[0163] ;

[0164] In the above formula, The increment of braking slip ratio The first derivative.

[0165] By unifying the incremental state-space equations under the two driving conditions and simplifying, the state-space equations for the longitudinal stability control strategy can be obtained as follows:

[0166] ;

[0167] In the above formula, The first derivative representing the increment of slip ratio; This represents the increment of the slip ratio; This represents the torque increment.

[0168] S3: Will As vehicle status x Then the state-space equations are transformed into standard form:

[0169] ;

[0170] In the above formula, Indicates vehicle state increment The first derivative; This indicates the control increment. ; A c The coefficients of the state and control terms are represented by . ; .

[0171] After discretization, it is represented as follows:

[0172] ;

[0173] Among them, △ x [ k ] and △ x [ k+ 1] respectively represent k and k Vehicle status at time +1; △ u [ k ]express k The torque increment at any given moment.

[0174] S4: Setting the cost function corresponding to the vehicle longitudinal stability control strategy based on the discretized state-space equations. J for:

[0175] ;

[0176] In the above formula, q 1 and q 2 represents the weighting coefficients for the status term and the control term, respectively; y [ k +j | k ]express k Time prediction k + j Vehicle status at any given time; y r [ k + j ]express k + j The target vehicle status at any given time; △ u [ k + j ]express k + j The torque increment at any given moment; p t Indicates the state step size; m t This indicates the control step size.

[0177] S5: Using any iterative optimization algorithm, based on the current state of the vehicle, iteratively optimize the cost function to achieve the desired result. J Minimize the torque increment at the next moment and use it to regulate the motor output torque of the vehicle.

[0178] In practical applications, when the cost function is minimized, the resulting series of motor output torque increments Δ from time k to time k+m-1 are obtained. u [ k ]~△ u [ k + m -1] can be represented as :

[0179]

[0180] For rear-wheel drive electric vehicles, the driving torque of their motors can be expressed as: T r :

[0181] ;

[0182] in, T tq This refers to the output torque of the motor. i 0 represents the reduction ratio; For mechanical transmission efficiency.

[0183] The first term that minimizes the cost function is selected as the control variable adjustment value, resulting in the optimized motor output torque. T tq It can be represented as:

[0184] .

[0185] In the above formula, for The vehicle's output torque at any given moment.

[0186] The output torque of this motor can be transmitted to the rear-wheel drive motor for control.

[0187] Example 3

[0188] Building upon the vehicle longitudinal stability control method based on optimal slip ratio provided in Embodiment 2, this embodiment further provides a vehicle longitudinal stability controller, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the vehicle longitudinal stability control method based on optimal slip ratio as described in Embodiment 2, and then performs real-time longitudinal stability control of the vehicle based on the collected vehicle wheel speed information and output torque.

[0189] In addition, this embodiment also provides a vehicle that employs the vehicle longitudinal stability controller as described above.

[0190] In a further optimized embodiment, the vehicle longitudinal stability controller also includes an extended state observer. It is assumed that when the front wheels act as driven wheels, their linear velocity can be approximated as the vehicle speed. However, in actual driving, dynamic conditions such as tire deformation may occur, and the dynamic model is a simplified model, which may introduce a certain degree of error. Therefore, an adaptive model predictive control method is designed based on the extended state observer, which improves the robustness of the control system while simultaneously estimating state variables and model disturbances.

[0191] In the extended state observer, the longitudinal dynamics model shows that the system is a first-order linear system, expressed as:

[0192] ;

[0193] In the above formula, For system state variables; For control input; For total disturbance; A is the system output; A and B are measurable system parameters.

[0194] Since the disturbance contains the derivative variable of the vehicle speed, the total disturbance can be extended into new state variables, resulting in an extended state vector. for The extended system equations are:

[0195] .

[0196] In the above formula, the coefficient matrix of each term Ae , B e , C e and E e They respectively satisfy: , , , .

[0197] The linear extended state observer is designed as follows:

[0198] ;

[0199] In the above formula, The estimated state for the extended state vector; This is an estimated value output by the system. L For the gain vector, .

[0200] The characteristic polynomial of the extended state observer is constructed by solving for the gain value using the pole placement method:

[0201] ;

[0202] in, s This represents the complex frequency variable in the Laplace transform.

[0203] Solving for the two gain terms using the coefficient matching method L 1. L 2:

[0204] ;

[0205] In the formula: This represents the observer bandwidth.

[0206] The state variables and disturbances obtained through discretization are as follows:

[0207] ;

[0208] In the above formula, for State variable estimates at time t; for k State variable estimates at time t; for The estimated value of the disturbance at time; for k The estimated value of the disturbance at time; for k The measured values ​​of the state variables at time t; The sampling interval is denoted as .

[0209] The estimated vehicle speed can be obtained from the estimated state variables, and the relationship between them is as follows:

[0210] ;

[0211] In the above formula, This is an estimated vehicle speed. for k The angular velocity of the rear wheel is obtained at any time.

[0212] like Figure 5 As shown, in the further optimized scheme including an extended state observer in this embodiment, the data sampler collects and processes wheel speed information, providing input to the slip ratio predictor through data processing. In the slip ratio predictor, the current real-time point information is compared with the stored operating condition map, and an arithmetic optimization algorithm is used to obtain the closest ratio curve, calculating the predicted value of the optimal slip ratio. The obtained optimal slip ratio value is then used as the control target input to the adaptive model predictive controller for control. Based on this, the incremental value of the control variable is calculated by minimizing the cost function, thereby calculating the optimized motor torque command and achieving longitudinal stability control of the vehicle.

[0213] Simulation Experiment

[0214] To verify the performance of the vehicle longitudinal stability control method based on the optimal slip ratio provided by this invention, technicians conducted simulations and tests on the relevant schemes.

[0215] In the simulation experiment, the weighting coefficient q 1. q 2 are 1 and Sampling interval T s The state step size is 0.001s. p t and control step size m t All are set to 1. The operating conditions are: braking from an initial speed of 80 km / h to 50 km / h, passing through connecting road sections with road adhesion ranging from low to high (0.8 to 1) and from high to low (1 to 0.5). The optimal slip ratios for road adhesion coefficients of 1, 0.8, and 0.5 are 0.1453, 0.1128, and 0.07, respectively.

[0216] In the simulation experiment, the vehicle's coefficient of adhesion, speed, slip ratio, and motor output torque changed over time as follows: Figure 6As shown in the figure, analysis of the data reveals that the present invention exhibits excellent controllability under both low-to-high and high-to-low adhesion coefficient variations. The estimated adhesion coefficient obtained by the extended state observer shows a high degree of similarity to the calculated adhesion coefficient. The front wheel linear velocity is highly similar to the vehicle speed, verifying the assumption that the front wheel linear velocity is approximately equal to the vehicle speed. Furthermore, the estimated vehicle speed obtained by the extended state observer also shows a high degree of similarity to the actual vehicle speed, verifying the accuracy of the model. The estimated slip ratio shows a high degree of similarity to the calculated real-time slip ratio, and the error between the predicted slip ratio and the actual optimal slip ratio is small, indicating high accuracy. The results demonstrate the adaptive performance of the extended state observer and the accuracy of the estimated values. Simultaneously, the optimal slip ratio prediction method also exhibits high accuracy.

[0217] Based on the data in the figure, the MSE (mean square error), RMSE (root mean square error), and MAPE (mean absolute percentage error) between the actual adhesion coefficient and the maximum adhesion coefficient are 0.0061197, 0.078229, and 2.8143%, respectively. The MSE, RMSE, and MAPE between the estimated adhesion coefficient and the maximum adhesion coefficient are 0.0060717, 0.077921, and 3.5569%, respectively. The MSE, RMSE, and MAPE between the real-time slip ratio and the estimated slip ratio are 2.299e-06, 0.0015162, and 1.2842%, respectively. The MSE, RMSE, and MAPE between the predicted optimal slip ratio and the optimal slip ratio are 8.199e-05, 0.0090549, and 3.9641%, respectively.

[0218] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time prediction method for optimal vehicle slip ratio based on ratio partitioning, characterized in that, It includes: Simultaneously collect vehicle adhesion coefficient under various road conditions With slip ratio Use it as road test data ( , Based on a large amount of road test data collected for each road condition, the corresponding slip curve is fitted, and the optimal slip ratio corresponding to each slip curve is determined. Furthermore, it obtains multiple slip curves of the current vehicle under different road conditions and their optimal slip ratio; The ratio of the real-time slip ratio to the optimal slip ratio in the slip curve is used as the slip ratio ratio, and then the slip ratio ratio of each data point in the slip curve is determined. The slip ratio ratio is divided into multiple gradient values ​​according to the preset accuracy level. The points with the same slip ratio on all slip curves are fitted to obtain the ratio curve of the corresponding gradient value. Real-time acquisition of vehicle wheel speed information, combined with the vehicle's dynamics model, generates the vehicle's adhesion coefficient in real time. and slip ratio This constitutes the current vehicle status. , ); Select the ratio curve that best matches the vehicle state from all ratio curves, and then calculate the real-time slip ratio. The ratio of the gradient value of the ratio curve is taken as the optimal slip ratio at the current moment.

2. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 1, characterized in that: Real-time adhesion coefficient of the vehicle The calculation formula is as follows: ； In the above formula, F zr This is the vertical force on the rear wheel; F xr This is the longitudinal force of the rear wheel; L This refers to the wheelbase between the front and rear wheels; G For vehicle weight; a This is the longitudinal distance from the vehicle's center of gravity to the front wheel; m For vehicle quality; h g The height of the vehicle's center of gravity; To accelerate the vehicle; This refers to the angular acceleration of the front wheel; r The radius of the wheel; I Let be the moment of inertia of the wheel.

3. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 2, characterized in that, Real-time vehicle slip rate The calculation formula is as follows: ； In the above formula, v For vehicle speed; This indicates the angular velocity of the rear wheel.

4. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 1, characterized in that: Multiple slip curves and their corresponding ratio curves under different road conditions are plotted on the same image as a working condition atlas; in the vehicle optimal slip rate update task, based on the vehicle state ( , ) Query the aforementioned operating condition graph, determine the closest ratio curve, and calculate the vehicle's current optimal slip ratio; And / or, The working condition map is dynamically updated. When the vehicle is driving under new road conditions, a new slip curve is fitted based on the real-time updated vehicle state value, and each ratio curve is dynamically updated based on the new slip curve.

5. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 4, characterized in that: The ratio curve that is closest to the current vehicle state is determined by calculating the minimum Euclidean distance between the current vehicle state value and each ratio curve in the working condition diagram. Alternatively, the slip curve and ratio curve in the working condition diagram can be represented in matrix form, and the ratio curve that is closest to the current vehicle state can be obtained through an arithmetic optimization algorithm.

6. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 5, characterized in that, Methods for finding the ratio curve that best matches the current vehicle state based on arithmetic optimization algorithms include: i. By fitting the slip curve with a polynomial function, then in the th... i On the slip curve, the ratio is p point ( , The following is represented: ； In the above formula, a i0 ~ a im The constant terms in the slip curve are respectively... m The coefficient of the second term; For the first i The optimal slip ratio on the slip curve; p This represents the ratio of real-time slip rate to the optimal slip rate. For the first i The ratio on the slip curve is p The point corresponds to the adhesion coefficient; ii. Let the number of attached sliding curves be n Different slip curves have the same ratio p Each point ( , ), ( , ), ..., ( , The following is represented: ； In the above formula, ; ; iii. For different slip curves with the same ratio p By fitting the data to each point, the corresponding ratio curve is obtained as follows: ； In the above formula, s 0~ s n-1 These are the constant terms in the ratio curve to... n The coefficient of the -1 term; iv. Different slip curves have the same ratio p The points are represented as follows: ； v. When any ratio curve is closest to the current vehicle state, the current vehicle's real-time utilization of the adhesion coefficient is... Corresponding to real-time slip rate The cost function is constructed by minimizing the difference in the adhesion coefficients of the ratio curves. D 1 is: ； vi. Preset ratio p The search range is determined by the ratio using an arithmetic optimization algorithm. p Optimize to minimize the cost function, and use this value as the ratio value of the ratio curve that is closest to the current vehicle state.

7. The real-time prediction method for optimal vehicle slip ratio based on ratio division as described in claim 6, characterized in that: Arithmetic optimization algorithm is used to compare ratios p The iterative optimization process includes: I. Initialize a uniformly distributed ratio using the following formula. p The initial population: ； In the above formula, Ub For ratio p The upper bound of the range of values; Lb For ratio p Lower bound of the value range; Rand Represents a random number between 0 and 1; The first in the population i The individual in the first j The location of 4-dimensional space; II. Use the Mathematical Function Accelerator (MOA) to determine whether to perform global exploration or local development. The MOA expression is: ； In the above formula, MOA( t () represents the current iteration number. t The acceleration function value; Min This represents the minimum value of the acceleration function; Max This represents the maximum value of the acceleration function; T This represents the maximum number of iterations. III. The probability of exploration and development is controlled using the mathematical optimizer (MOP). The calculation formula is as follows: ； In the above formula, MOP( t ) represents the current mathematical optimizer probability; For iteration sensitivity coefficient, A higher value results in a higher number of iterations for MOP ( t The greater the impact, the stronger the influence. IV. When random numbers r If 1 < MOA, proceed to the exploration phase; otherwise, proceed to the development phase. During the exploration phase, based on random numbers r 2. Decide whether to use a multiplication or division strategy to update the population. The update formula is: ； In the above formula, r 2 is a random number between 0 and 1; X( t +1) represents the updated position of the individual; X b ( t () represents the position of the individual with the best fitness before the update; For the control coefficient of the search process; To adjust for the minimum value of the probability; During the development phase, based on random numbers r 3. Decide whether to use an addition or subtraction strategy to update the population. The update formula is: ； V. Iteratively update the population and calculate the cost function value of each individual in the population in each round as its fitness value. Select the individual with the smallest fitness value as the optimal individual in each round. When the preset maximum number of iterations is reached, output the optimal individual as the optimization result.

8. A vehicle longitudinal stability control method based on optimal slip ratio, characterized in that, It includes the following steps: S1: The optimal slip ratio of the vehicle is generated in real time using the real-time prediction method of vehicle optimal slip ratio based on ratio division as described in any one of claims 1-7. ; S2: Using the optimal slip ratio from the previous step as the target slip ratio, construct the state-space equations for the vehicle longitudinal stability control strategy as follows: ； In the above formula, The first derivative representing the increment of slip ratio; This represents the increment of the slip ratio; For torque increment; S3: Will As vehicle status x Then the state-space equation is transformed into standard form: ； In the above formula, Indicates vehicle state increment The first derivative; This indicates the control increment. ; A c The coefficients of the state and control terms are represented by . ; ; After discretization, it is represented as follows: ； Among them, △ x [ k ] and △ x [ k+ 1] respectively represent k and k Vehicle status at time +1; △ u [ k ]express k The torque increment at any given moment; S4: Setting the cost function corresponding to the vehicle longitudinal stability control strategy based on the discretized state-space equations. J for: ； In the above formula, q 1 and q 2 represents the weighting coefficients for the status term and the control term, respectively; y [ k + j | k ]express k Time prediction k + j Vehicle status at any given time; y r [ k + j ]express k + j The target vehicle status at any given time; △ u [ k + j ]express k + j The torque increment at any given moment; p t Indicates the state step size; m t Indicates the control step size; S5: Using any iterative optimization algorithm, based on the current state of the vehicle, iteratively optimize the cost function to achieve the desired result. J Minimize the torque increment at the next moment and use it to regulate the motor output torque of the vehicle.

9. A vehicle longitudinal stability controller, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the vehicle longitudinal stability control method based on the optimal slip ratio as described in claim 8, and then performs longitudinal stability control on the vehicle in real time based on the collected vehicle wheel speed information and output torque.

10. A vehicle, characterized in that: It employs the vehicle longitudinal stability controller as described in claim 9.

Images