A method and system for dynamic selection of vehicle braking wheels driven by stability risk field

By using a stability risk field-driven approach to dynamically select braking wheels, the shortcomings of existing wheel selection strategies under multi-objective conflict conditions are addressed. This approach achieves a balance between dynamic adjustment of vehicle stability control and response speed, ensuring braking performance.

CN123061333APending Publication Date: 2026-11-10JILIN UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611395202.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-09
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing vehicle stability braking control methods cannot continuously and dynamically adjust wheel selection strategies according to vehicle status under multi-objective conflict conditions. This makes it difficult for fixed threshold rules to adapt to changes in load, road surface adhesion, and roll state, leading to conflicts in stability control objectives.

Method used

A stability risk field-driven approach is adopted. By acquiring the vehicle state vector, the yaw rate, sideslip angle and their allowable boundaries are calculated to construct a stability risk coordinate vector. Combined with the risk field gradient and sensitivity vector, the braking wheels are dynamically selected to construct a comprehensive wheel selection scoring system, realize forward-looking dynamic selection, and introduce a signal credibility classification and degradation mechanism.

Benefits of technology

It achieves dynamic migration of brake wheel selection, taking into account both response speed and multi-objective coordination capabilities, avoiding erroneous interventions caused by unreliable data, and ensuring braking performance under low adhesion or high load transfer conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123061333A_ABST
    Figure CN123061333A_ABST
Patent Text Reader

Abstract

This invention discloses a method and system for dynamic selection of vehicle braking wheels driven by a stability risk field. The method acquires the vehicle state vector, calculates the reference yaw rate, center of gravity sideslip angle and boundary conditions, as well as rollover, adhesion, and wheel load risks, constructing a risk coordinate vector. A weighted sum of squares is used as the risk field value, and the gradient is calculated. Candidate wheels are generated based on steering and rollover. A reference braking torque is applied to each candidate wheel to predict risk changes, obtaining a sensitivity vector, and the desired yaw moment is calculated. The predicted risk reduction is calculated from the gradient and sensitivity. A comprehensive score is constructed by combining indicators such as yaw consistency, rollover suppression, wheel load availability, adhesion saturation, and target conflict. A single wheel or a main wheel plus compensation wheel set is selected based on the score. Pressure output is allocated based on the comprehensive reliability classification and downgrading. This invention achieves unified characterization of multi-objective risks and dynamic wheel optimization, improving stability and control reliability under complex operating conditions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the fields of automotive chassis electronic control and vehicle active safety technology, specifically to a method and system for dynamic selection of vehicle braking wheels driven by a stability risk field. Background Technology

[0002] When vehicles are engaged in high-speed cornering, emergency obstacle avoidance, cornering braking, or driving on low-traction surfaces, stability issues may arise simultaneously due to limited tire working boundaries, load transfer, and the coupling of yaw and roll movements. These issues include insufficient yaw response, excessive yaw response, increased sideslip angle, increased load on the outer wheels, and unloading of the inner wheels. These problems are particularly pronounced in vehicles with a high center of gravity and a wide range of load variations (such as SUVs, MPVs, and commercial vehicles). Stability control systems typically generate additional yaw moments through differential braking or reduce vehicle speed and lateral acceleration through braking the outer wheels to suppress yaw instability and rollover risks.

[0003] Existing vehicle stability braking control methods typically trigger brake wheel selection based on front and rear axle yaw rates, yaw deviation, lateral acceleration thresholds, or rollover thresholds. While this approach is easy to calibrate under conditions dominated by a single risk, fixed inner and outer wheel rules can lead to conflicting control objectives when yaw stability and rollover prevention requirements coexist. For example, braking a particular wheel might generate the desired yaw moment but could increase outer load transfer or exacerbate tire adhesion utilization; conversely, braking an outer wheel might reduce rollover risk but introduce additional yaw moment contrary to yaw stability requirements.

[0004] For vehicles with a high center of gravity and a wide range of load variations, relying solely on fixed thresholds to determine whether the front outer wheel, front inner wheel, rear outer wheel, or rear inner wheel should brake is insufficient to continuously adjust the wheel selection strategy in response to changes in load weight, road surface adhesion, roll state, wheel load state, and tire force margin. Therefore, it is necessary to propose a dynamic wheel selection method that inherits the candidate braking rules for both inner and outer wheels while incorporating stability risk field identification, risk gradient calculation, and wheel-end action prediction. This allows the braking wheel selection to shift in real time based on the vehicle's current risk distribution and the control actions of each wheel. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for dynamic selection of vehicle braking wheels driven by a stability risk field, so as to solve the technical problem that existing fixed threshold rules cannot continuously and dynamically adjust the wheel selection strategy according to the vehicle state under multi-objective conflict conditions.

[0006] To achieve the above objectives, the technical solution provided by this invention is: a method for dynamic selection of vehicle braking wheels driven by a stability risk field, comprising the following steps: S1: Obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. S2: The weighted sum of squares of each component of the stability risk coordinate vector is used as the stability risk field value; the partial derivative of the stability risk field value with respect to each risk coordinate component is obtained to obtain the risk field gradient; the inner and outer wheels are determined according to the steering direction, and a set of candidate braking wheels is generated based on the steering state and rollover risk level. S3: Apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. S4: Calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. Combine the yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score for each wheel. Select the candidate braking wheel with the highest comprehensive wheel selection score as the main braking wheel. Based on the single wheel feasibility conditions and the combination of the remaining candidate braking wheels and the main braking wheel, predict the risk reduction and determine whether to use the main braking wheel alone or add a compensation wheel to obtain the final set of braking wheels. S5: Calculate the overall credibility of the risk field; based on the comparison between the overall credibility and the preset threshold, perform graded downgrading, adjust the weight of each risk coordinate component in the stability risk field value, constrain the member range of the candidate braking wheel set, and limit the maximum braking pressure output; allocate braking pressure and output it based on the finally selected single wheel or wheel group and its overall wheel selection score.

[0007] To optimize the above technical solution, the specific measures also include: In step S1, the vehicle state vector includes a vehicle planar motion state sub-vector, a rollover and adhesion state sub-vector, and a four-wheel vertical load sub-vector; the reference yaw rate and the reference center of mass sideslip angle and their allowable boundaries are calculated based on a two-degree-of-freedom vehicle reference model; the rollover risk value is calculated based on the lateral load transfer rate, which is the ratio of the difference between the vertical loads of the left and right wheels of the vehicle to the sum of the vertical loads; the tire adhesion utilization rate is determined according to the ratio of the tire force of each wheel to the tire adhesion limit, and the tire adhesion utilization rate risk value is determined according to the degree to which the tire adhesion utilization rate exceeds the preset adhesion risk boundary; the wheel load deficiency risk value is determined according to the degree to which the vertical load of each wheel is insufficient relative to the minimum allowable vertical load.

[0008] Furthermore, in step S2, the stability risk field value is a weighted sum of the squares of yaw error, sideslip error, rollover risk, tire adhesion utilization risk, and wheel load insufficiency risk. The weight of each risk coordinate component is adaptively adjusted according to the non-negative intensity of the corresponding risk component. When the non-negative intensity of a certain risk component exceeds the preset threshold for reinforcement, the weight of that risk component increases with the increase of risk intensity until it reaches the upper limit of the weight corresponding to the preset high-risk threshold.

[0009] In step S2, determining the inner and outer wheels based on the steering direction specifically means: when the equivalent front wheel average turning angle is greater than the preset steering dead zone threshold, the vehicle turns in the positive direction, with the right wheel being the outer wheel and the left wheel being the inner wheel; when the equivalent front wheel average turning angle is less than the preset steering dead zone threshold negative value, the vehicle turns in the negative direction, with the left wheel being the outer wheel and the right wheel being the inner wheel. Furthermore, when the vehicle is in an understeer state, the inner wheels of the front and rear axles are included in the candidate braking wheel set; when the vehicle is in an oversteer state, the outer wheels of the front axle are included in the candidate braking wheel set; when the rollover risk increases, the outer front wheel for turning is added; when the rollover risk enters the high-risk zone and the outer rear wheel meets both the wheel load access condition and the attachment access condition, the outer rear wheel is added.

[0010] In step S3, the comprehensive expected additional yaw moment is composed of a weighted average of the yaw stabilization additional moment requirement and the anti-rollover correction moment requirement; the yaw stabilization additional moment requirement is determined by the yaw angular velocity error and the center of gravity sideslip angle error; the direction of the anti-rollover correction moment requirement is opposite to the current turning yaw direction, and its amplitude is determined by the rollover risk value; wherein the weighting coefficient changes continuously according to the rollover risk value, when the rollover risk is zero, the yaw stabilization additional moment requirement is used entirely, and when the rollover risk is at its maximum value, the anti-rollover correction moment requirement is used entirely.

[0011] In step S4, the predicted risk reduction is the negative value of the dot product of the risk field gradient and the sensitivity vector; the yaw direction consistency is a measure of the consistency between the direction of the additional yaw moment generated by the candidate braking wheel and the direction of the comprehensive expected additional yaw moment; the rollover suppression benefit is the predicted reduction in the rollover risk value after braking of the candidate braking wheel; the wheel load availability benefit is determined based on the margin of the predicted vertical load of the candidate braking wheel relative to the minimum allowable vertical load; and the adhesion saturation penalty is determined based on the degree to which the tire adhesion utilization rate of the candidate braking wheel exceeds a preset saturation threshold.

[0012] Further, in step S4, the specific determination method for wheel selection is as follows: First, it is determined whether the wheel with the highest comprehensive wheel selection score meets the preset single-wheel feasibility conditions. The single-wheel feasibility conditions include the risk reduction exceeding the significant reduction threshold, the comprehensive score not lower than the minimum score threshold, the target conflict penalty not exceeding the maximum allowable conflict value, and the yaw direction consistency not lower than the minimum consistency requirement. When the single-wheel feasibility conditions are met and the difference between the highest score and the second highest score is greater than the single-wheel selection difference threshold, the single-wheel mode is selected. Otherwise, when the rollover risk value and the yaw error value simultaneously exceed their respective high-risk boundaries, the wheel group mode is activated. The wheel group consists of the outer main braking wheel and the dynamically selected compensation wheel.

[0013] In step S5, based on the comparison result between the comprehensive confidence level and the preset threshold, a graded downgrade is performed. Specifically, the comprehensive confidence level of the risk field is determined by the minimum value of the vehicle state signal confidence level, wheel load estimation confidence level, adhesion estimation confidence level, center of gravity sideslip angle estimation confidence level, roll angle estimation confidence level, and brake actuator confidence level. When the comprehensive confidence level is higher than the high confidence level threshold, the risk weights identified online and the complete candidate brake wheel set are used. When the comprehensive confidence level is between the low confidence level threshold and the high confidence level threshold, the risk weights of the previous effective period are frozen and the candidate brake wheel set is restricted to safe access wheels. When the comprehensive confidence level is lower than the low confidence level threshold, the pre-calibrated downgraded risk weights are used, and the candidate brake wheel set is restricted to the intersection of the basic inner and outer wheel candidate sets and the safe access wheel set, while the maximum output braking pressure is limited.

[0014] Further, in step S5, the method of allocating braking pressure based on the comprehensive wheel selection score specifically involves: normalizing the comprehensive wheel selection score of the final braking wheel after taking a non-negative value to obtain the braking pressure allocation coefficient of each wheel; and using the product of the braking pressure allocation coefficient and the total additional braking pressure as the target additional braking pressure of each wheel. When the final braking wheel is a single wheel, the allocation coefficient of that wheel is one; when the final braking wheel is a wheel set, the allocation coefficient of each wheel is proportional to the non-negative value of its comprehensive wheel selection score; and the allocation coefficient of unselected braking wheels is zero.

[0015] As another important technical solution, the present invention also provides a vehicle braking wheel dynamic selection system driven by a stability risk field, realizing the vehicle braking wheel dynamic selection method driven by a stability risk field as described above, including: The risk coordinate construction module is used to obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. The candidate wheel generation module is used to take the weighted sum of squares of each component of the stability risk coordinate vector as the stability risk field value; to obtain the risk field gradient by taking the partial derivative of each risk coordinate; to determine the inner and outer wheels according to the steering direction; and to generate a set of candidate braking wheels according to the steering state and rollover risk level. The sensitivity prediction and expected torque calculation module is used to apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; and calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. The wheel selection decision module is used to calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. It combines yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score. If the wheel with the highest score meets the preset single-wheel feasibility conditions, it is selected as the final braking wheel. Otherwise, the main braking wheel is determined from the candidate braking wheel set, and the compensation wheel is dynamically selected to pair with it based on the combined risk reduction, wheel load adhesion admission conditions and yaw moment compensation effective conditions to form the final braking wheel set. The output and adjustment module is used to allocate and output braking pressure based on the comprehensive wheel selection score of a single wheel or wheel set; calculate the comprehensive confidence level of the risk field, freeze the current risk weight and restrict the safe access of candidate wheels when the confidence level is medium, switch to conservative weight and return to the basic candidate logic when the confidence level is low, and limit the maximum braking pressure.

[0016] The present invention also proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements a dynamic selection method for vehicle braking wheels driven by a stability risk field as described above.

[0017] The present invention also proposes a computer-readable storage medium storing a computer program that causes a computer to execute a vehicle braking wheel dynamic selection method driven by a stability risk field as described above.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention integrates five stability indices—yaw, lateral slip, rollover, adhesion, and wheel load—into a risk field framework. Through adaptive weight adjustment, the control focus naturally shifts with changes in risk, thus solving the problem of difficulty in coordinating multiple conflicting objectives. This allows the selection of brake wheels to no longer rely on fixed inner and outer rules, but to dynamically migrate with changes in loading, adhesion, and roll.

[0019] This invention predicts the risk reduction effect of each wheel through pre-braking and constructs a three-dimensional scoring system by combining factors such as yaw consistency, rollover suppression, wheel load availability, adhesion saturation and target conflict, thus realizing the forward-looking dynamic selection of wheels; at the same time, it automatically switches between two modes, single-wheel rapid response and wheel group coordinated braking, according to the risk level, taking into account both response speed and multi-target coordination capability.

[0020] This invention introduces a graded degradation mechanism based on signal, estimation, and actuator reliability. When information quality deteriorates, it adopts gradual measures such as freezing weights, limiting candidate wheels, or switching conservative parameters to avoid erroneous intervention due to unreliable data. This effectively ensures the actual execution effect of braking commands under low adhesion or high load transfer conditions. Attached Figure Description

[0021] Figure 1 : A schematic diagram of the process in an embodiment of the present invention. Detailed Implementation

[0022] The present invention will be further described in detail below through specific embodiments, but it should not be construed as limiting the scope of the subject matter of the present invention to the following embodiments. All technologies implemented based on the above content of the present invention fall within the scope of the present invention.

[0023] This invention provides a method for dynamic selection of vehicle braking wheels driven by a stability risk field, comprising the following steps: S1: Obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. Read the vehicle state vector in each control cycle. x ( k ), including the vehicle's planar motion state sub-vectors x m ( k Side rollover and attachment state sub-vectors x s ( k ) and four-wheel vertical load sub-vector x w ( k This data is used for subsequent risk field identification, generation of basic candidate braking wheels, and prediction of wheel-end braking effects. Specifically:

[0024]

[0025]

[0026]

[0027] in, v x ( k The vehicle's longitudinal velocity is obtained by fusing the four wheel speeds and longitudinal acceleration through a Kalman filter. d ( k () represents the equivalent average front wheel steering angle; r ( k () represents the actual yaw rate; β ( k () represents the actual centroid sideslip angle; a x ( k )and a y ( k These are the longitudinal acceleration and the lateral acceleration, respectively. f ( k )and ( k ) represent the vehicle roll angle and roll rate, respectively, estimated by a roll state observer driven by lateral acceleration and roll rate; LTR( k ) represents the lateral load transfer rate; m ( k () represents the current road surface adhesion coefficient; F z,i ( k ) indicates the first i Estimated vertical load on each wheel. i =FL,FR,RL,RR represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; the superscript T indicates transpose.

[0028] In some implementations, preset basic vehicle parameters, including wheelbase, are read. L Distance from the center of gravity to the front and rear axles l f and l r Front and rear track width t f and t r Effective rolling radius of the wheel R w Current vehicle quality m ( k ), center of mass height h g yaw moment of inertia I z And roll-related parameters; the equivalent average front wheel steering angle is calculated from the steering wheel angle according to the current steering gear ratio: ; in, d sw ( k () indicates the steering wheel angle. i s ( k () indicates the steering gear ratio. The equivalent average front wheel steering angle is calculated accordingly.

[0029] The sideslip angle of the center of mass is estimated by a discrete-state observer, which uses the lateral velocity and yaw rate of the center of mass as states and the yaw rate and lateral acceleration as measurements. The sideslip angle is calculated from the lateral velocity of the center of mass estimated by the observer, specifically as follows:

[0030] in, This represents the lateral velocity of the center of mass estimated by the state observer; This indicates a preset positive number for low-speed protection. This indicates that the larger value is taken to avoid slow division by zero.

[0031] The vertical load on each of the four wheels is estimated wheel by wheel based on the static axle load, longitudinal and lateral acceleration, current vehicle mass, center of gravity height, front and rear track width, roll angle, and roll rate, according to a pre-calibrated load transfer model.

[0032] in, F z,i 0 ( k ) indicates the first i The static vertical load of each wheel; G x,i and G y,i These represent the distribution coefficient of longitudinal load transfer to the wheel and the direction coefficient of lateral load transfer to the wheel, respectively, and their signs are predetermined according to the vehicle coordinate system and the wheel position. Indicates the front and rear axle roll moment distribution coefficient; Indicates the first i The wheelbase of each wheel on the axle; K φ,i and C φ,i These represent the equivalent roll stiffness and roll damping allocated to the axle, respectively.

[0033] In some implementations, the roll angle is output by a roll state observer, and the actual wheel-end braking pressure is obtained from feedback by pressure sensors in the electronic braking system. The longitudinal tire force is estimated based on the pressure-braking torque calibration relationship and the wheel rotation equation, while the lateral tire force is estimated by a four-wheel force observer constrained by tire adhesion.

[0034] All wheels that can be independently braked Oh for:

[0035] Where FL, FR, RL, and RR represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. Subsequent attachment utilization and wheel load risk are... Oh Internal calculations show that the candidate set of braking wheels satisfies... Oh c ( k ) Oh The final brake wheel assembly meets the requirements. Oh ( k ) Oh c ( k ).

[0036] The reference yaw rate and reference centroid sideslip angle, along with their allowable boundaries, are calculated based on a two-degree-of-freedom vehicle reference model, specifically as follows: Reference yaw rate r ref ( k The calculation formula is:

[0037] in, The understeer coefficient is calculated from the current vehicle weight, wheelbase, and equivalent lateral stiffness of the front and rear axles. The equivalent lateral stiffness of the front and rear axles is obtained from a pre-calibrated two-dimensional table. The table lookup input is the current road adhesion coefficient and the vertical loads of the front and rear axles.

[0038] Reference centroid sideslip angle β ref ( k The formula for calculating its allowable boundary is as follows:

[0039]

[0040] in, This represents the safety factor of the allowable boundary of the centroid sideslip angle, and is not greater than one. Indicates the reference centroid sideslip angle boundary; , and These represent the reference road surface adhesion coefficient, the reference vehicle mass, and the reference vehicle speed, respectively. and These represent the preset lower and upper limits of the centroid sideslip angle boundary, respectively; This indicates that the calculation result will be restricted between the lower and upper limits; current quantity , and Obtained from the vehicle control system.

[0041] Rollover risk value R ρ ( k Based on the lateral load transfer rate (LTR) k )calculate:

[0042]

[0043] Among them, LT R 1 indicates that the risk of rollover is beginning to increase at the boundary, LT R 2 represents the high rollover risk boundary, satisfying 0 <LT R 1 <LT R 2≤1, neither of them is LTR ( k The component of ) and in the fixed-boundary embodiment, it does not have a time subscript; sat [0,1] (·) indicates a saturated function that restricts the result to between zero and one; F z,R ( k )and F z,L ( k ) represent the sum of the vertical loads on the right and left wheels, respectively.

[0044] In some implementations, tire adhesion utilization risk value u μ ( k The ratio of the current tire force to the tire adhesion limit for each wheel is as follows:

[0045]

[0046] or μ ( k )=max i∈Ω or μ,i ( k ); m μ ( k )=1- or μ ( k ) in, or μ ( k This indicates the maximum adhesion utilization rate among the four wheels; or μ,1and or μ,2 These represent the boundary where the attachment risk begins to increase and the high-risk boundary, respectively. F x,i ( k )and F y,i ( k ) respectively represent the longitudinal direction of the vehicle coordinate system x Axial and transverse y The first in the axial direction i The force of each wheel tire; or μ,i ( k ) indicates the first i The adhesion utilization rate of each wheel; m μ ( k The value indicates the minimum adhesion margin of the entire vehicle; the smaller the value, the closer it is to the adhesion limit.

[0047] Insufficient wheel load risk value u F ( k This is a risk metric when the ratio of the current vertical load to the static vertical load of each wheel is lower than a preset threshold, specifically:

[0048] in, F z,min This indicates the minimum vertical load required for effective wheel braking.

[0049] Constructing a stability risk coordinate vector z ( k ):

[0050]

[0051]

[0052] in, z r ( k )and z β ( k ) represent the normalized yaw error and the normalized lateral deviation error, respectively; This indicates the normalized scale of the yaw error; due to R ρ ( k Using a saturation function, when |LTR( k )| <LT R When the rollover risk coordinate is 1, it is zero, and there will be no false rollover risk due to the square of a negative value.

[0053] S2: The weighted sum of squares of each component of the stability risk coordinate vector is used as the stability risk field value; the partial derivative of the stability risk field value with respect to each risk coordinate component is obtained to obtain the risk field gradient; the inner and outer wheels are determined according to the steering direction, and a set of candidate braking wheels is generated based on the steering state and rollover risk level. In some implementations, the absolute values ​​of the normalized yaw error and sideslip error are taken respectively, and together with the rollover, adhesion, and wheel load risk values, they are used as the non-negative intensity of the corresponding risk. For each type of risk, its weight is adaptively adjusted with the risk intensity according to the following formula:

[0054] in, An index representing five types of risks: yaw, side slip, rollover, adhesion, and wheel load; Indicates the first The non-negative intensity of the risk class is taken in sequence. , , , and ; Indicates the basic risk weight; Indicates the amount of weight enhancement; and These respectively represent the boundary where this type of risk begins to strengthen and the boundary where it becomes high-risk; This indicates a positive number to prevent the denominator from being too small.

[0055] The stability risk field value is:

[0056] in, α r ( k ), α β ( k ), α ρ ( k ), α μ ( k )and α F ( k The values ​​represent the risk weights for yaw, side slip, rollover, adhesion, and wheel load, respectively. These weights are adaptively adjusted based on the non-negative intensity of the corresponding risk component. When the non-negative intensity of a certain risk component exceeds the preset threshold for increasing intensity, the weight of that risk component increases with the increase in risk intensity until it reaches the upper limit of the weight corresponding to the preset high-risk threshold.

[0057]

[0058] in, Represents the risk field relative to the risk coordinate vector The gradient column vector, This indicates that the gradient of the risk coordinate vector is calculated; the risk weights are updated first in each control cycle, and the risk weights are treated as constants during the gradient calculation in this cycle; This represents the negative gradient row vector, i.e., the descent direction of the risk field.

[0059] In some implementations, the default inner and outer wheels are determined based on the steering direction, and the mapping of the inner and outer sides of the front and rear axles is checked using the equivalent sideslip angle sign of the front and rear axles, specifically: When the steering input is greater than the positive dead zone threshold, the right wheel is the default outer wheel and the left wheel is the default inner wheel; when the steering input is less than the negative dead zone threshold, the left wheel is the default outer wheel and the right wheel is the default inner wheel; when the steering input is within the dead zone, the previous effective mapping is maintained, and the equivalent sideslip angles of the front and rear axles are used to check the inner and outer mappings of the corresponding axles.

[0060]

[0061]

[0062] in, and These represent the set of front axle wheels and the set of rear axle wheels, respectively. and These represent the outer wheels of the front axle and the outer wheels of the rear axle, respectively. and These represent the inner wheels of the corresponding axles; and These represent the equivalent tire slip angles for the front and rear axles, respectively.

[0063] In some implementations, after determining the inner and outer wheels of the front and rear axles, a basic candidate braking wheel set is generated based on the vehicle's yaw stability state. Specifically, when the vehicle is in an understeer state, the inner wheels of the front and rear axles are included in the basic candidate braking wheel set; when the vehicle is in an oversteer state, the outer wheels of the front axle are included in the basic candidate braking wheel set. Based on this, the candidate braking wheel set is expanded according to rollover risk: when the rollover risk increases, the outer wheels of the front axle are added; when the rollover risk enters a high-risk zone, and the outer wheels of the rear axle simultaneously meet the wheel load access condition and the adhesion access condition, the outer wheels of the rear axle are added.

[0064] in, This represents the set of basic candidate wheels obtained by filtering from the basic candidate vectors; This refers to the set of outer wheels added based on the risk of rollover; The union of the two is limited to the set of all controllable wheels. The final set of candidate brake wheels obtained within.

[0065] S3: Apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. Apply the same reference braking torque increment to each wheel in the candidate braking wheel set. DM b,ref To make the risk reduction capabilities of different wheels comparable, predictions are made by first calculating the longitudinal braking force and additional yaw moment at the wheel end, and then obtaining the changes in each risk coordinate component after braking from the vehicle prediction model.

[0066]

[0067] in, ΔF x,i ( k ) indicates the first i The longitudinal braking force increment generated by each wheel under the reference braking torque increment; D M z,i ( k The sign indicates the corresponding additional yaw moment. y i This indicates the lateral lever arm of the wheel relative to the vehicle's center of gravity; the left and right wheels have opposite signs.

[0068]

[0069] in, Indicates only to the first Increment of reference braking torque applied to each wheel Then, the risk coordinates are obtained by the vehicle single-step prediction model based on planar motion, roll, wheel load transfer and tire force constraints, and by risk coordinate mapping in step S1. Indicates the current risk coordinates before braking; This indicates the change in the risk coordinates of the wheel before and after reference braking.

[0070]

[0071] in, B i ( k ) indicates the first iThe sensitivity vector of the effect of wheel braking on the risk coordinate has five components, which correspond to yaw error, sideslip error, rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, respectively. M b,i This indicates the additional braking torque applied to the corresponding wheel; the controller follows... B i ( k )=( Δz i ( k )) / ( DM b,ref )≈( z ( k )) / ( M b,i The same reference braking torque increment is used for finite difference online calculation. The vector is updated periodically based on the current vehicle mass, center of gravity height, estimated vertical load of the four wheels, and current road adhesion coefficient.

[0072] The overall expected additional yaw moment is a weighted average of the additional yaw stabilization moment requirement and the anti-rollover correction moment requirement:

[0073]

[0074]

[0075] in, This indicates the additional torque required for yaw stability; This indicates the anti-rollover correction torque requirement, which is in the opposite direction to the current yaw direction during the turn. , and Indicates the corresponding feedback gain; Represents a symbolic function; This represents the rollover risk participation weight obtained by continuous mapping of rollover risk values, with a value range of zero to one. When hour, ;when At that time, the two types of demand are continuously weighted; when hour, This embodiment directly takes... .

[0076] S4: Calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. Combine the yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score for each wheel. Select the candidate braking wheel with the highest comprehensive wheel selection score as the main braking wheel. Based on the single wheel feasibility conditions and the combination of the remaining candidate braking wheels and the main braking wheel, predict the risk reduction and determine whether to use the main braking wheel alone or add a compensation wheel to obtain the final set of braking wheels. The method for calculating the predicted risk reduction is as follows:

[0077] in, Indicates the first The first-order predicted risk reduction of a candidate wheel is positive, indicating that the wheel causes the vehicle state to move in the downward direction of the risk field. express Risk reduction returns after normalization to a preset scale and limitation to between zero and one.

[0078] In some implementations, the yaw direction determination measure is the product of the overall expected additional yaw moment and the corresponding wheel's predicted additional yaw moment; the yaw direction consistency is mapped to zero, half, or one according to the following formula:

[0079] in, Indicates the first The yaw direction discrimination of each wheel; This indicates a sign function with a dead zone, when the input is greater than... Take one when the input is less than 1. When the absolute value of the input is not greater than 1, take the negative one. Take zero at time; This represents the dead zone threshold of the function; It indicates the consistency of the lateral direction. When the directions are consistent, it takes one; when the directions are opposite, it takes zero; and when it is in the dead zone, it takes half.

[0080] The comprehensive score also includes rollover mitigation benefits, wheel load availability benefits, and adhesion saturation penalties. Among these, This indicates that rollover reduces returns, which is due to the reduction in rollover risk. According to the normalization scale Normalization yields, for The rollover risk component in the middle; This represents the wheel load availability benefit, calculated from the current vertical load margin relative to the minimum permissible wheel load, based on a reference value. Normalization yields the result; This indicates an adhesion saturation penalty, which occurs when the adhesion utilization rate... Exceeding the threshold for taking effect It then increases, and is based on the maximum allowable value. Limitation; all the above quantities are normalized to between zero and one.

[0081] In some implementations, the target conflict penalty P C,i ( k )for:

[0082] in, Indicates the first Penalty for conflicting objectives on each wheel; This represents the normalized scale of the risk reduction. The risk reduction return is calculated using... calculate.

[0083] Overall wheel selection score for each wheel S i ( k The calculation method is as follows:

[0084]

[0085]

[0086]

[0087] in, S i ( k The higher the value, the more suitable the wheel is to be selected; w D , w H , w ρ , w L , w μ and w C All weights are non-negative.

[0088] In some implementations, candidate wheels are ranked according to a comprehensive wheel selection score. When the wheel with the highest score simultaneously meets the conditions of risk reduction, lowest score, maximum permissible conflict, and minimum directional consistency, and the difference between its score and the second highest score wheel reaches a preset threshold, single-wheel braking is applied; when the rollover risk and yaw risk are both high and there is obvious conflict on a single wheel, the wheel set selection is initiated.

[0089] Among them, the candidate wheels with the highest and second highest scores are obtained by ranking the comprehensive wheel selection scores; when the candidate set contains only one wheel, the second highest score is defined as negative infinity; the risk reduction threshold, minimum score threshold, maximum allowable conflict, minimum directional consistency and high risk threshold used for single wheel feasibility conditions and wheel set requirements are all pre-calibrated values.

[0090] In wheel set mode, the outer wheel with the best overall performance in risk reduction, rollover suppression, wheel load availability and adhesion constraint is selected from the candidate set as the main braking wheel. Then, the wheel that simultaneously meets the conditions of wheel load and adhesion access, effective reduction of combined risk and effective compensation of yaw moment is selected from the remaining candidate wheels. The wheel with the highest combined score is selected as the compensation wheel.

[0091]

[0092]

[0093]

[0094] in, Indicates the final set of braking wheels; Indicates the selected wheel in single-wheel mode; and These represent the main brake wheel and compensation wheel selected in wheel set mode, respectively. and These represent the single-wheel mode flag and the wheel set mode flag, respectively. In single-wheel mode, the set contains only one wheel, while in wheel set mode, the set contains two wheels. When any of the following three conditions occur: the maximum candidate risk return is not positive, the attachment access condition is not met, or the wheel load access condition is not met, no new stability braking pressure will be added in this cycle.

[0095]

[0096] in, S i + ( k () indicates the non-negative portion of the score; G S ( k () represents the sum of the non-negative parts of the selected wheel scores; c i ( k The symbol represents the brake pressure distribution coefficient. When only one wheel is selected, the distribution coefficient for that wheel is set to one; the distribution coefficient for unselected wheels is set to zero.

[0097]

[0098] in, P b,i( k ) indicates the first i The target additional braking pressure for each wheel; P b ( k This represents the total additional braking pressure required for stability control. ΔP inc and ΔP dec These represent the single-cycle pressure increase and pressure decrease limits, respectively. When the anti-lock braking system (ABS) intervenes, the target additional pressure on the corresponding wheel is limited to the upper limit allowed by the ABS controller; when the wheel slip ratio exceeds the threshold, pressure is decreased according to the single-cycle pressure decrease limit; when the actuator reports a limitation, the corresponding wheel is removed from the candidate set, and its stability additional pressure is gradually reduced to zero according to the single-cycle pressure decrease limit.

[0099] S5: Calculate the overall credibility of the risk field; based on the comparison between the overall credibility and the preset threshold, perform graded downgrading, adjust the weight of each risk coordinate component in the stability risk field value, constrain the member range of the candidate braking wheel set, and limit the maximum braking pressure output; allocate braking pressure and output it based on the finally selected single wheel or wheel group and its overall wheel selection score.

[0100] In some implementations, the credibility of each input and estimate is calculated based on sensor self-diagnosis, data timeliness, observation residuals, model consistency, and brake actuator tracking error, and then fused to obtain the overall credibility of the risk field, specifically as follows:

[0101] in, The confidence index represents six categories: vehicle status, wheel load, road surface adhesion, center of gravity lateral slip, vehicle body roll, and brake actuator. Indicates the first The reliability of the input signal or estimator; The corresponding residuals are represented, where the acquired signals use sensor consistency residuals, the state estimates use observation residuals, and the braking actuators use pressure tracking errors. Indicates the maximum allowable residual; when any of the following anomalies is detected: hardware self-diagnostic failure, data timeout, or data out of bounds, the corresponding confidence level is set to zero. Indicates the first The reliability of the individual brake actuator; This represents the minimum reliability of all controllable wheel brake actuators; This indicates the overall credibility of the risk field. When the above anomalies do not occur, it is based on... Calculation; Actuator reliability according to calculate.

[0102]

[0103] in, q H and q L These represent the high and low confidence thresholds, respectively. α j,cons This represents a pre-set fixed risk weight under low confidence conditions. Under high confidence conditions, an online risk weight is used; under medium confidence conditions, the previous effective weight is frozen; and under low confidence conditions, the candidate for risk gradient expansion is removed, and the system returns to the basic inner and outer wheel candidate logic, retaining only wheels that meet the safety admission criteria. This represents the j-th type of risk weight currently used for control.

[0104]

[0105] in, This represents the set of candidate braking wheels adjusted based on the overall credibility of the risk field. Indicates the first Safety access mark for each wheel: only when , and If both conditions are met, take 1; otherwise, take 0. This represents a conservative wheel load threshold, which is not less than the aforementioned minimum wheel load threshold. This represents a conservative adhesion utilization threshold, which is less than the aforementioned maximum allowable adhesion utilization threshold. This represents the set of wheels for which all safety access signs are selected as one; This represents the intersection of the basic inner and outer candidate sets and the safe set. At medium confidence, candidate rounds are restricted; at low confidence, the risk gradient expansion candidate set is exited, and the conservative basic logic is returned.

[0106]

[0107]

[0108]

[0109] in, This indicates the reference value for the adhesion margin; This represents the minimum value among risk field confidence, actuator confidence, normalized attachment margin, and candidate wheel minimum load availability, when... Time Agreement ; Indicates the maximum allowable additional braking pressure of the system; Indicates Revised maximum allowable additional braking pressure; Indicates output up to the th The target additional braking pressure of each brake actuator is addressed by implementing graded degradation through risk field freezing, conservative weight replacement, candidate wheel limitation, and brake output limiting, rather than directly shutting down all stability controls when the confidence of the state estimate decreases.

[0110] In another embodiment, the present invention proposes a vehicle braking wheel dynamic selection system driven by a stability risk field, realizing the vehicle braking wheel dynamic selection method driven by a stability risk field as described above, including: The risk coordinate construction module is used to obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. The candidate wheel generation module is used to take the weighted sum of squares of each component of the stability risk coordinate vector as the stability risk field value; to obtain the risk field gradient by taking the partial derivative of each risk coordinate; to determine the inner and outer wheels according to the steering direction; and to generate a set of candidate braking wheels according to the steering state and rollover risk level. The sensitivity prediction and expected torque calculation module is used to apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; and calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. The wheel selection decision module is used to calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. It combines yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score. If the wheel with the highest score meets the preset single-wheel feasibility conditions, it is selected as the final braking wheel. Otherwise, the main braking wheel is determined from the candidate braking wheel set, and the compensation wheel is dynamically selected to pair with it based on the combined risk reduction, wheel load adhesion admission conditions and yaw moment compensation effective conditions to form the final braking wheel set. The output and adjustment module is used to allocate and output braking pressure based on the comprehensive wheel selection score of a single wheel or wheel set; calculate the comprehensive confidence level of the risk field, freeze the current risk weight and restrict the safe access of candidate wheels when the confidence level is medium, switch to conservative weight and return to the basic candidate logic when the confidence level is low, and limit the maximum braking pressure.

[0111] In another embodiment of the present invention, an electronic device is proposed, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements a dynamic selection method for vehicle braking wheels driven by a stability risk field as described above.

[0112] In another embodiment of the present invention, a computer-readable storage medium is provided storing a computer program that causes a computer to execute a vehicle braking wheel dynamic selection method driven by a stability risk field as described above.

[0113] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0114] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent substitutions, and improvements made by those skilled in the art to the above embodiments without departing from the scope of the technical solution of the present invention, based on the technical essence of the present invention, shall still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for dynamic selection of vehicle braking wheels driven by a stability risk field, characterized in that, Includes the following steps: S1: Obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. S2: The weighted sum of squares of each component of the stability risk coordinate vector is taken as the stability risk field value; The partial derivative of the stability risk field value with respect to each risk coordinate component is obtained to obtain the risk field gradient; the inner and outer wheels are determined according to the steering direction, and a set of candidate braking wheels is generated based on the steering state and rollover risk level; S3: Apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. S4: Calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. Combine the yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score for each wheel. Select the candidate braking wheel with the highest comprehensive wheel selection score as the main braking wheel. Based on the single wheel feasibility conditions and the combination of the remaining candidate braking wheels and the main braking wheel, predict the risk reduction and determine whether to use the main braking wheel alone or add a compensation wheel to obtain the final set of braking wheels. S5: Calculate the overall credibility of the risk field; based on the comparison between the overall credibility and the preset threshold, perform graded downgrading, adjust the weight of each risk coordinate component in the stability risk field value, constrain the member range of the candidate braking wheel set, and limit the maximum braking pressure output; allocate braking pressure and output it based on the finally selected single wheel or wheel group and its overall wheel selection score.

2. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S1, the vehicle state vector includes a vehicle planar motion state sub-vector, a rollover and adhesion state sub-vector, and a four-wheel vertical load sub-vector; the reference yaw rate and the reference center of mass sideslip angle and their allowable boundaries are calculated based on a two-degree-of-freedom vehicle reference model; the rollover risk value is calculated based on the lateral load transfer rate, which is the ratio of the difference between the vertical loads of the left and right wheels of the vehicle to the sum of the vertical loads; the tire adhesion utilization rate is determined according to the ratio of the tire force of each wheel to the tire adhesion limit, and the tire adhesion utilization rate risk value is determined according to the degree to which the tire adhesion utilization rate exceeds the preset adhesion risk boundary; the wheel load deficiency risk value is determined according to the degree to which the vertical load of each wheel is insufficient relative to the minimum allowable vertical load.

3. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S2, the risk field gradient is obtained by taking the first-order partial derivative of the stability risk field value with respect to each risk coordinate component; the candidate braking wheel set is generated as follows: the outer front axle wheel, inner front axle wheel, outer rear axle wheel, and inner rear axle wheel are determined according to the steering direction; when the vehicle is in an understeer state, the inner front and rear axle wheels are included in the candidate braking wheel set; when the vehicle is in an oversteer state, the outer front axle wheel is included in the candidate braking wheel set; when the rollover risk increases, the outer front wheel is added; when the rollover risk enters the high-risk zone and the outer rear wheel simultaneously meets the wheel load access condition and the attachment access condition, the outer rear wheel is added.

4. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S3, the comprehensive expected additional yaw moment is composed of a weighted average of the yaw stabilization additional moment requirement and the anti-rollover correction moment requirement; the yaw stabilization additional moment requirement is determined by the yaw angular velocity error and the center of gravity sideslip angle error; the direction of the anti-rollover correction moment requirement is opposite to the current turning yaw direction, and its amplitude is determined by the rollover risk value; wherein the weighting coefficient changes continuously according to the rollover risk value, when the rollover risk is zero, the yaw stabilization additional moment requirement is used entirely, and when the rollover risk is at its maximum value, the anti-rollover correction moment requirement is used entirely.

5. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S4, the predicted risk reduction is the negative value of the dot product of the risk field gradient and the sensitivity vector; the yaw direction consistency is a measure of the consistency between the direction of the additional yaw moment generated by the candidate braking wheel and the direction of the comprehensive expected additional yaw moment; the rollover suppression benefit is the predicted reduction in the rollover risk value after braking of the candidate braking wheel; the wheel load availability benefit is determined based on the margin of the predicted vertical load of the candidate braking wheel relative to the minimum allowable vertical load; and the adhesion saturation penalty is determined based on the degree to which the tire adhesion utilization rate of the candidate braking wheel exceeds a preset saturation threshold.

6. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S5, based on the comparison result between the comprehensive confidence level and the preset threshold, a graded downgrade is performed. Specifically, the comprehensive confidence level of the risk field is determined by the minimum value of the vehicle state signal confidence level, wheel load estimation confidence level, adhesion estimation confidence level, center of gravity sideslip angle estimation confidence level, roll angle estimation confidence level, and brake actuator confidence level. When the comprehensive confidence level is higher than the high confidence level threshold, the risk weights identified online and the complete candidate brake wheel set are used. When the comprehensive confidence level is between the low confidence level threshold and the high confidence level threshold, the risk weights of the previous effective period are frozen and the candidate brake wheel set is restricted to safe access wheels. When the comprehensive confidence level is lower than the low confidence level threshold, the preset fixed risk weights are used, and the candidate brake wheel set is returned to the basic inner and outer wheel candidate logic, retaining only the wheels that meet the safe access conditions, while limiting the maximum output braking pressure.

7. The method for dynamic selection of vehicle braking wheels driven by a stability risk field according to claim 1, characterized in that: In step S5, the method of allocating braking pressure based on the comprehensive wheel selection score is as follows: after taking the comprehensive wheel selection score of the final braking wheel as a non-negative value and normalizing it, the braking pressure allocation coefficient of each wheel is obtained, and the product of the braking pressure allocation coefficient and the total additional braking pressure is used as the target additional braking pressure of each wheel; when the final braking wheel is a single wheel, the allocation coefficient of that wheel is one. When the final braking wheel is a wheel set, the distribution coefficient of each wheel is proportional to the non-negative value of its comprehensive wheel selection score; the distribution coefficient of the unselected braking wheel is zero.

8. A vehicle braking wheel dynamic selection system driven by a stability risk field, used to implement the vehicle braking wheel dynamic selection method driven by a stability risk field as described in any one of claims 1 to 7, characterized in that, include: The risk coordinate construction module is used to obtain the vehicle state vector, calculate the reference yaw rate, reference centroid sideslip angle and their allowable boundaries, calculate the yaw error and sideslip error based on the actual values ​​and reference values, calculate the rollover risk value, tire adhesion utilization risk value and wheel load underload risk value, and construct the stability risk coordinate vector. The candidate round generation module is used to take the weighted sum of squares of each component of the stability risk coordinate vector as the stability risk field value; The partial derivative of the stability risk field value with respect to each risk coordinate component is obtained to obtain the risk field gradient; the inner and outer wheels are determined according to the steering direction, and a set of candidate braking wheels is generated based on the steering state and rollover risk level; The sensitivity prediction and expected torque calculation module is used to apply the same reference braking torque increment to each wheel in the candidate braking wheel set, predict the state change after braking and calculate the change of each risk coordinate component to obtain the sensitivity vector of each wheel to the risk coordinate; and calculate the comprehensive expected additional yaw moment based on yaw error, sideslip error and rollover risk. The wheel selection decision module is used to calculate the predicted risk reduction based on the risk field gradient and sensitivity vector. It combines yaw direction consistency, rollover suppression benefit, wheel load availability benefit, adhesion saturation penalty and target conflict penalty to construct a comprehensive wheel selection score for each wheel. The candidate braking wheel with the highest comprehensive wheel selection score is selected as the main braking wheel. Based on the feasibility of a single wheel and the combination of the remaining candidate braking wheels and the main braking wheel, the module predicts the risk reduction and determines whether to use the main braking wheel alone or add a compensation wheel to obtain the final set of braking wheels. The output and adjustment module is used to calculate the overall credibility of the risk field; based on the comparison result of the overall credibility and the preset threshold, it performs graded downgrading, adjusts the weight of each risk coordinate component in the stability risk field value, constrains the member range of the candidate braking wheel set, and limits the maximum braking pressure output; and allocates and outputs the braking pressure according to the finally selected single wheel or wheel group and its overall wheel selection score.

9. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements a dynamic selection method for vehicle braking wheels driven by a stability risk field as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program causes the computer to execute a dynamic selection method for vehicle braking wheels driven by a stability risk field as described in any one of claims 1 to 7.