A method, device, medium and product for identifying road surface adhesion conditions

Abstract

The application discloses a road surface adhesion condition recognition method, device, medium and product, and relates to the field of vehicle electronic braking. The method comprises the following steps: taking the road surface longitudinal slip stiffness as a state quantity, taking the vehicle longitudinal deceleration as an observation quantity, calculating an observation residual by using a Kalman filtering algorithm, and performing sliding window statistics on the observation residual to obtain a residual sequence covariance average value; two judgment quantities are constructed according to the residual sequence covariance average value, and an adaptive fading factor is determined according to the comparison result of the judgment quantities; the prior error covariance matrix is corrected according to the adaptive fading factor to obtain a corrected prior error covariance matrix; the road surface adhesion condition representation quantity is calculated according to the corrected prior error covariance matrix, the observation residual and the current time prior stiffness estimation value; and the vehicle road surface adhesion condition is determined according to the road surface adhesion condition representation quantity. The technical scheme of the embodiment of the application can accurately recognize the adhesion condition of the road surface for the vehicle in real time.

Description

Technical Field

[0001] This invention relates to the field of vehicle electronic braking technology, and in particular to a method, device, medium, and product for identifying road surface adhesion conditions. Background Technology

[0002] Real-time identification of road surface adhesion conditions is fundamental for chassis electronic control algorithms such as anti-lock braking systems, electronic brake force distribution, and stability control to achieve good control performance. "Road surface adhesion conditions" do not involve a precise estimation of the road surface adhesion coefficient, but rather an identification of whether the current road surface is classified as high, medium, or low adhesion.

[0003] Due to the variable operating environment, large curb weight, and significant changes in center of gravity, higher demands are placed on the real-time performance and accuracy of road surface category recognition. During vehicle start-up or braking, wheel speed signals experience significant noise due to sensor resolution and vehicle vibration. Furthermore, false high slip rates are detected during non-effective braking phases, leading to misclassification of normal road surfaces as low-adhesion surfaces. Traditional single-wheel independent recognition algorithms struggle to distinguish between genuine low-adhesion responses and transient disturbances during the initial pressure build-up phase, easily resulting in distorted recognition results. Moreover, existing traditional recursive least squares methods or standard Kalman filters exhibit small error covariance matrices after reaching steady state, resulting in higher weighting of model predictions compared to lower weighting of current observations. Consequently, recognition results tend to lag when the vehicle enters a connecting road surface or when road surface adhesion conditions change abruptly. Summary of the Invention

[0004] This invention provides a method, device, medium, and product for identifying road surface adhesion conditions, in order to solve the problems of poor accuracy and lag in existing road surface adhesion condition identification methods.

[0005] According to one aspect of the present invention, a method for identifying road surface adhesion conditions is provided, comprising: Using the longitudinal slip stiffness of the road surface as the state variable and the longitudinal deceleration of the vehicle as the observation, the observation residuals are calculated using the Kalman filter algorithm, and the observation residuals are statistically analyzed using a sliding window to obtain the average covariance of the residual sequence. Based on the mean covariance of the residual sequence, the covariance matrix of the observation noise, the longitudinal slip rate of the wheel, and the covariance matrix of the prior error, two decision variables are constructed, and the adaptive fading factor is determined based on the comparison results of the decision variables. Based on the adaptive fading factor, the prior error covariance matrix is ​​corrected to obtain the corrected prior error covariance matrix; Based on the corrected prior error covariance matrix, observation residuals, and the current time-time prior stiffness estimate, the road surface adhesion condition characterization quantity is calculated. Based on the road surface adhesion condition characterization quantity, the first road surface adhesion judgment threshold, and the second road surface adhesion judgment threshold, the front axle and rear axle adhesion category judgment results are determined, and the whole vehicle road surface adhesion conditions are determined based on the front axle and rear axle adhesion category judgment results.

[0006] According to another aspect of the present invention, a vehicle-mounted device is provided, the vehicle-mounted device comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the road surface adhesion condition identification method according to any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the method for identifying road surface adhesion conditions according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the method for identifying road surface adhesion conditions as described in any embodiment of the present invention.

[0009] The technical solution of this invention uses road longitudinal slip stiffness as the state variable and vehicle longitudinal deceleration as the observation. It employs a Kalman filter algorithm to calculate the observation residuals and performs sliding window statistics on the residuals to obtain the average covariance of the residual sequence. Based on the average covariance of the residual sequence, the observation noise covariance matrix, the wheel longitudinal slip rate, and the prior error covariance matrix, two decision variables are constructed. An adaptive fading factor is determined based on the comparison results of the decision variables. Then, the prior error covariance matrix is ​​corrected based on the adaptive fading factor to obtain a corrected prior error covariance matrix. Further, based on the corrected prior error covariance matrix, the observation residuals, and the current time-based prior stiffness estimate, a road surface adhesion condition characterization variable is calculated. Based on the road surface adhesion condition characterization variable, a first road surface condition adhesion decision threshold, and a second road surface condition adhesion decision threshold, the front and rear axle adhesion category determination results are determined. Finally, based on the front and rear axle adhesion category determination results, the overall vehicle road surface adhesion conditions are determined. This solution requires no additional sensor hardware. By adjusting the adaptive fading factor logic, it compares the theoretical residual with the actual residual in real time, weakening the historical inertia of the old road surface condition and thus shortening the recognition convergence time under sudden changes in road surface adhesion conditions. Furthermore, by utilizing the physical constraints of axle load transfer, it logically filters the slip artifacts caused by the reduction of load on the rear axle, ensuring that the rear axle will not prematurely reduce pressure due to misidentification when braking on a good road surface, thereby ensuring braking strength and achieving highly robust online estimation of road surface conditions. This solves the problems of poor accuracy and recognition lag in existing road surface adhesion condition recognition, and can accurately identify the vehicle's adhesion conditions to the road surface in real time.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating a method for identifying road surface adhesion conditions provided in Embodiment 1 of the present invention; Figure 2 A schematic diagram of the execution logic of a commercial vehicle road surface adhesion condition recognition device; Figure 3 A schematic diagram of the structure of an in-vehicle device that can be used to implement an embodiment of the present invention is shown. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0015] Example 1 Figure 1 This is a flowchart of a method for identifying road surface adhesion conditions according to Embodiment 1 of the present invention. This embodiment is applicable to the case of online accurate identification of road surface adhesion conditions. The method can be executed by a road surface adhesion condition identification device (a virtual device that performs the road surface adhesion condition identification method). The road surface adhesion condition identification device can be implemented in hardware and / or software, and can be configured in an on-board device. Figure 1 As shown, the method includes: Step 110: Using the longitudinal slip stiffness of the road surface as the state variable and the longitudinal deceleration of the vehicle as the observation, the observation residuals are calculated using the Kalman filter algorithm, and the observation residuals are statistically analyzed using a sliding window to obtain the average covariance of the residual sequence.

[0016] Among these, the longitudinal slip stiffness of the road surface can be used to describe the ability of the road surface to generate longitudinal forces when in contact with the tire. The observed residuals can be used to characterize the degree of mismatch between the measured longitudinal deceleration of the vehicle and the longitudinal deceleration predicted based on the old state. The average covariance of the residual sequence can be the average of the covariance of the observed residuals.

[0017] In this embodiment of the invention, the longitudinal slip stiffness of the road surface can be used as the state variable and the longitudinal deceleration of the vehicle as the observation. Using the Kalman filter algorithm, the observation residual is calculated in each sampling period (i.e., the sampling period of the original vehicle motion data). Then, the observation residual is statistically analyzed by sliding window to calculate the average covariance of the residual sequence.

[0018] In an optional embodiment of the present invention, before calculating the observation residual using the Kalman filter algorithm with the road surface longitudinal slip stiffness as the state variable and the vehicle longitudinal deceleration as the observation, the method may further include: correcting the wheel speed according to the vehicle yaw angle to obtain the wheel speed compensation value; and calculating the wheel longitudinal slip ratio according to the wheel speed compensation value and the vehicle reference speed.

[0019] The wheel speed compensation value can be a correction result of the wheel speed based on the vehicle yaw angle. The vehicle reference speed can be the real-time vehicle speed. The wheel longitudinal slip ratio can be used as a kinematic quantity to describe the degree of tire slippage.

[0020] In this embodiment of the invention, the wheel speed compensation value can be calculated based on the product of the original wheel angular velocity measurement value and the effective rolling radius of the tire, the vehicle yaw rate, and the vehicle track width. Then, the difference between the vehicle reference speed and the wheel speed compensation value is divided by the sum of the vehicle reference speed and a small constant to obtain the wheel longitudinal slip ratio.

[0021] Among them, the small constant is used to prevent the denominator from being zero.

[0022] Step 120: Construct two decision variables based on the mean covariance of the residual sequence, the covariance matrix of the observation noise, the longitudinal slip rate of the wheel, and the covariance matrix of the prior error, and determine the adaptive fading factor based on the comparison results of the decision variables.

[0023] The observation noise covariance matrix describes the correlation between the magnitude of sensor measurement error fluctuations and each measurement dimension. The prior error covariance matrix characterizes the uncertainty of state estimation during the prediction phase. The decision variable can be an intermediate variable in calculating the adaptive fading factor. The adaptive fading factor can be a coefficient used to correct the prediction covariance update process.

[0024] In this embodiment of the invention, a decision quantity can be constructed based on the mean covariance of the residual sequence and the covariance matrix of the observation noise, and another decision quantity can be constructed based on the longitudinal slip ratio of the wheel and the covariance matrix of the prior error. Then, matrix trace operation is performed on the two decision quantities, and the adaptive fading factor corresponding to the comparison result of the matrix trace operation results of the two decision quantities is determined.

[0025] In an optional embodiment of the present invention, two decision quantities are constructed based on the mean covariance of the residual sequence, the covariance matrix of the observation noise, the longitudinal slip ratio of the wheel, and the covariance matrix of the prior error. This may include: using the difference between the mean covariance of the residual sequence and the covariance matrix of the observation noise as the first decision quantity; and determining the second decision quantity based on the longitudinal slip ratio of the wheel and the covariance matrix of the prior error.

[0026] Among them, the first decision factor and the second decision factor can be two decision factors for determining the adaptive fading factor.

[0027] In this embodiment of the invention, the difference between the average covariance of the residual sequence and the covariance matrix of the observation noise can be used as the first criterion, and the wheel longitudinal slip ratio, the prior error covariance matrix, and the transpose of the wheel longitudinal slip ratio can be calculated by matrix multiplication to obtain the second criterion.

[0028] Step 130: Based on the adaptive fading factor, correct the prior error covariance matrix to obtain the corrected prior error covariance matrix.

[0029] Among them, the corrected prior error covariance matrix can be the prior error covariance matrix corrected based on the adaptive fading factor.

[0030] In this embodiment of the invention, the prior error covariance matrix of the Kalman filter algorithm when calculating the Kalman gain at the current time can be corrected based on the adaptive fading factor as the weight coefficient, so as to obtain the corrected prior error covariance matrix.

[0031] In an optional embodiment of the present invention, correcting the prior error covariance matrix according to the adaptive fading factor to obtain the corrected prior error covariance matrix may include: correcting the prior error covariance matrix based on the following formula: ;in, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. The process noise covariance matrix is... As an adaptive fading factor, To correct the prior error covariance matrix.

[0032] The posterior error covariance matrix and the state transition matrix can be calculated based on the Kalman filter algorithm. The process noise covariance matrix can be used to characterize the statistical uncertainty caused by the slow changes in road surface conditions and model approximation errors, and can be preset according to the system's dynamic response characteristics, historical estimation error statistics, or empirical calibration parameters.

[0033] Step 140: Calculate the road surface adhesion condition characterization quantity based on the corrected prior error covariance matrix, observation residuals, and the prior stiffness estimate at the current time.

[0034] The prior stiffness estimate at the current moment can be the prior estimate of the longitudinal slip stiffness of the road surface at the current moment (referred to as the prior stiffness estimate at the current moment). The road surface adhesion condition characterization quantity can be the posterior estimate of the longitudinal slip stiffness of the road surface at the current moment after filtering correction.

[0035] In this embodiment of the invention, the Kalman filter gain can be calculated based on the corrected prior error covariance matrix according to the Kalman filter algorithm. The calculated Kalman filter gain is then multiplied by the observation residual at the current time, and the product value is summed with the corrected prior error covariance matrix to obtain the road surface adhesion condition characterization quantity.

[0036] In an optional embodiment of the present invention, calculating the road surface adhesion condition characterization quantity based on the corrected prior error covariance matrix, the observation residual, and the prior stiffness estimate at the current time may include: calculating the Kalman gain based on the corrected prior error covariance matrix, the wheel longitudinal slip ratio, and the observation noise covariance matrix; and summing the product of the Kalman gain and the observation residual at the current time with the prior stiffness estimate at the current time to obtain the road surface adhesion condition characterization quantity.

[0037] In this embodiment of the invention, the Kalman gain at the current moment can be calculated according to the corrected prior error covariance matrix, wheel longitudinal slip ratio, and observation noise covariance matrix corresponding to the current moment, using the Kalman filtering algorithm. The product of the Kalman gain and the observation residual at the current moment is then summed with the prior stiffness estimate at the current moment to serve as the road surface adhesion condition characterization quantity, thus allowing the road surface adhesion condition characterization quantity of the front and rear axle wheels to be calculated.

[0038] Step 150: Determine the front axle and rear axle adhesion category determination results based on the road surface adhesion condition characterization quantity, the first road surface adhesion determination threshold, and the second road surface adhesion determination threshold, and determine the whole vehicle road surface adhesion conditions based on the front axle and rear axle adhesion category determination results.

[0039] The first road surface condition adhesion determination threshold can be used to determine the threshold for high adhesion types. The second road surface condition adhesion determination threshold can be used to determine the threshold for low adhesion types. The front axle and rear axle adhesion category determination results can be the identification results of the road surface adhesion types of the front axle and rear axle. The whole vehicle road surface adhesion condition can be the finally determined adhesion type of the whole vehicle to the current road surface. The whole vehicle road surface adhesion condition can include the adhesion type of the vehicle's front axle to the ground and the adhesion type of the vehicle's rear axle to the ground.

[0040] In this embodiment of the invention, when the road surface adhesion condition characterization quantity is greater than the first road surface adhesion judgment threshold, the road surface adhesion condition of the wheel can be determined to be of the high adhesion type. When the road surface adhesion condition characterization quantity is less than the second road surface adhesion judgment threshold, the road surface adhesion condition of the wheel can be determined to be of the low adhesion type. When the road surface adhesion condition characterization quantity is greater than or equal to the second road surface adhesion judgment threshold and less than or equal to the first road surface adhesion judgment threshold, the road surface adhesion condition of the wheel is determined to be of the medium adhesion type. Therefore, based on the above-mentioned road surface adhesion condition judgment rules of the wheel and the road surface adhesion condition characterization quantity of the front and rear axle wheels, the front and rear axle adhesion category judgment results can be determined. Then, when the front and rear axle adhesion category judgment results are consistent, the road surface adhesion condition of the whole vehicle can be obtained.

[0041] In an optional embodiment of the present invention, determining the front axle and rear axle adhesion category determination results based on the road surface adhesion condition characterization quantity, the first road surface adhesion determination threshold, and the second road surface adhesion determination threshold may include: when the vehicle state meets the calculation conditions for the road surface adhesion condition characterization quantity, determining a score correction value based on the range within which the road surface adhesion condition characterization quantity falls; summing the score correction value with the cumulative road surface condition score of the previous time to obtain an initial road surface condition discrimination score; and determining the front axle and rear axle adhesion category determination results based on the initial road surface condition discrimination score, the first road surface adhesion determination threshold, and the second road surface adhesion determination threshold.

[0042] The calculation conditions for the road surface adhesion condition characterization quantity can be pre-set conditions for activating the scoring of the road surface adhesion condition characterization quantity. Optionally, the calculation conditions for the road surface adhesion condition characterization quantity can include a vehicle reference speed higher than a set speed threshold, brake air pressure higher than a set brake trigger threshold, a stable pressure change gradient, and wheel longitudinal slip ratio falling within a reliable estimation range. The scoring correction value can be a value that has a mapping relationship with the numerical range into which the road surface adhesion condition characterization quantity falls. For example, when the road surface adhesion condition characterization quantity is located in a first preset data range (corresponding to the numerical range of high adhesion type), the scoring correction value is... (Positive bonus weight), that is, the first preset data interval and A mapping relationship exists. When the road surface adhesion condition characterization value is within the second preset data interval, the scoring correction value is a default minimum positive value or 0, meaning the second preset data interval has a mapping relationship with the default minimum positive value / 0. When the road surface adhesion condition characterization value is within the third preset data interval, the scoring correction value is... (Negative deduction weight), that is, the third preset data interval and There is a mapping relationship. The cumulative pavement condition score can be the pavement condition score calculated from the previous time step, with the initial cumulative pavement condition score being 0. The initial pavement condition discrimination score can be the sum of the score correction value and the cumulative pavement condition score.

[0043] Correspondingly, when the vehicle state meets the calculation conditions for the road surface adhesion condition representation quantity, the range within which the road surface adhesion condition representation quantity falls can be determined. This leads to the determination of the score correction value mapped to the range of the road surface adhesion condition representation quantity. The score correction value is then summed with the cumulative road surface condition score from the previous moment to obtain the initial road surface condition discrimination score. Following the wheel's road surface adhesion condition judgment rules, this initial discrimination score is compared with the first and second road surface condition adhesion judgment thresholds to obtain the front and rear axle adhesion category judgment results. Through an integral mechanism, features are extracted during stable braking phases, avoiding recognition jumps during unsteady vehicle processes.

[0044] In an optional embodiment of the present invention, determining the vehicle road surface adhesion conditions based on the front axle and rear axle adhesion category determination results may include: when the front axle and rear axle adhesion category determination results are inconsistent, and the front axle adhesion category is the first adhesion type and the rear axle adhesion type is the second adhesion type, if it is determined that there is pseudo-slippage in the rear axle dynamic load based on the rear axle real-time normal load and the rear axle static load, then the rear axle adhesion category is corrected to the first adhesion type.

[0045] The first adhesion type can be a high adhesion type. The second adhesion type can be a low adhesion type. The real-time normal load on the rear axle can be used to describe the actual load level on the rear axle during vehicle braking. The static load on the rear axle can be used to describe the load level on the vehicle's rear axle under reference conditions.

[0046] In this embodiment of the invention, if the front axle and rear axle attachment category determination results are inconsistent, and the front axle attachment category is the first attachment type while the rear axle attachment type is the second attachment type, the decrease ratio of the real-time normal load of the rear axle to the static load of the rear axle can be calculated. If the decrease ratio is greater than a set criterion, the rear axle is determined to be in a light-load state, and it is further determined that the rear axle identification deviation at this time is a transient disturbance pseudo-slip phenomenon caused by the reduction of axle load, and the rear axle attachment category is corrected to the first attachment type. If the decrease ratio is less than or equal to the set criterion, the rear axle attachment type does not need to be corrected. When the front axle and rear axle attachment category determination results are inconsistent, and the operating condition is not that the front axle attachment type is the first attachment type while the rear axle attachment type is the second attachment type, the front axle and rear axle attachment category determination results can be maintained, or the correction can be made according to the decrease ratio of the real-time normal load of the rear axle to the static load of the rear axle.

[0047] The technical solution of this invention uses road longitudinal slip stiffness as the state variable and vehicle longitudinal deceleration as the observation. It employs a Kalman filter algorithm to calculate the observation residuals and performs sliding window statistics on the residuals to obtain the average covariance of the residual sequence. Based on the average covariance of the residual sequence, the observation noise covariance matrix, the wheel longitudinal slip rate, and the prior error covariance matrix, two decision variables are constructed. An adaptive fading factor is determined based on the comparison results of the decision variables. Then, the prior error covariance matrix is ​​corrected based on the adaptive fading factor to obtain a corrected prior error covariance matrix. Further, based on the corrected prior error covariance matrix, the observation residuals, and the current time-based prior stiffness estimate, a road surface adhesion condition characterization variable is calculated. Based on the road surface adhesion condition characterization variable, a first road surface condition adhesion decision threshold, and a second road surface condition adhesion decision threshold, the front and rear axle adhesion category determination results are determined. Finally, based on the front and rear axle adhesion category determination results, the overall vehicle road surface adhesion conditions are determined. This solution requires no additional sensor hardware. By adjusting the adaptive fading factor logic, it compares the theoretical residual with the actual residual in real time, weakening the historical inertia of the old road surface condition and thus shortening the recognition convergence time under sudden changes in road surface adhesion conditions. Furthermore, by utilizing the physical constraints of axle load transfer, it logically filters the slip artifacts caused by the reduction of load on the rear axle, ensuring that the rear axle will not prematurely reduce pressure due to misidentification when braking on a good road surface, thereby ensuring braking strength and achieving highly robust online estimation of road surface conditions. This solves the problems of poor accuracy and recognition lag in existing road surface adhesion condition recognition, and can accurately identify the vehicle's adhesion conditions to the road surface in real time.

[0048] Example 2 This invention provides a specific example of a method for identifying road surface adhesion conditions in Embodiment 2. Technical terms that are the same as or correspond to those in the above embodiments will not be repeated here.

[0049] In a specific example, the commercial vehicle road surface adhesion condition recognition device comprises four sequentially coupled functional modules: a signal acquisition and kinematic decoupling compensation module, an adaptive fading Kalman filter online estimation module, a multidimensional feature confidence integration module, and a dynamic axle load collaborative correction module. Each module executes sequentially according to the hierarchy of "signal preprocessing, physical quantity estimation, category determination, and physical consistency correction." The execution logic of the commercial vehicle road surface adhesion condition recognition device can be found in [reference needed]. Figure 2 .

[0050] The signal acquisition and kinematic decoupling compensation module is used to acquire raw signals during the control cycle. Under steering and braking conditions, the radii of the inner and outer wheel travel paths are different, which will cause the slip ratio calculation to include geometric deviations. Therefore, the yaw rate is used to correct the left and right wheel speeds.

[0051] The calculation method for the wheel speed compensation value of the left wheel is as follows: ; The calculation method for the wheel speed compensation value of the right wheel is as follows: .

[0052] in, This is the circumferential linear velocity of the left wheel after compensation, i.e., the left wheel speed compensation value; This is the circumferential linear velocity of the right wheel after compensation, i.e., the right wheel speed compensation value; This is the original angular velocity measurement value of the left wheel; This is the original angular velocity measurement value of the right wheel; The effective rolling radius of the tire; The vehicle's yaw rate; This refers to the vehicle's wheelbase.

[0053] The longitudinal slip ratio of the wheel is: ; In the formula, This refers to the longitudinal slip ratio of the wheel; For reference vehicle speed; To standardize the wheel speed notation after independent correction for each wheel, corresponding to the following for left and right wheel scenarios respectively. and ; To prevent tiny constants with a denominator of zero.

[0054] An online estimation module for adaptive fading Kalman filtering is used for online estimation of longitudinal slip stiffness based on multi-window residual adaptive fading Kalman filtering.

[0055] This module uses the vehicle's longitudinal dynamic response to infer road surface physical characteristics, but it does not directly output the road adhesion coefficient. Instead, it first estimates an intermediate physical quantity online to characterize the current road adhesion conditions: the road longitudinal slip stiffness. This quantity represents the slope of the tire longitudinal force as a function of the slip ratio within the linear region during the initial braking phase. Its dynamic changes are closely related to the road adhesion conditions, and therefore can serve as a core feature for subsequent adhesion category identification. It should be noted that the basic recursive process of standard Kalman filtering includes the following steps: Let the longitudinal slip stiffness of the road surface be the state variable, and the longitudinal deceleration of the vehicle be the observed variable. Within each sampling period, the prior estimate of the current time is first obtained from the posterior estimate of the previous time step. Then, the deceleration prediction value is obtained from the prior estimate and the current slip ratio. The residual is obtained by subtracting the predicted value from the measured value. Subsequently, the Kalman gain is calculated based on the prediction covariance, and the state is updated using the Kalman gain and the residual to obtain the posterior estimate of the current time step.

[0056] ; .

[0057] in, for Time-prior stiffness estimate. for The posterior stiffness estimate at time step 1 corresponds to the filtered and corrected output of the previous sampling period. for The longitudinal deceleration of the vehicle is measured at any given moment, i.e., the measurement observed at the current moment. for The longitudinal slip ratio of the wheel is calculated in real time. This is process noise; For observation noise. Among them, Instead of directly entering the stiffness output result, it first forms the observation residual together with the deceleration prediction value, and then participates in the subsequent state update through the residual. and These are used to describe model uncertainties in the state evolution process and noise disturbances in the observation and measurement process, respectively. They do not directly solve for their instantaneous values, but rather use the noise covariance matrix from subsequent processes. Covariance matrix of observation noise The statistical form reflects its impact.

[0058] The observation residual at the current time is calculated based on the following formula: ; In the formula, This represents the observation residual at the current moment. Initial stiffness values ​​can be pre-defined at the algorithm's inception. The algorithm then iteratively updates the initial error covariance based on the sampling period.

[0059] In standard Kalman filtering, the Kalman gain at the current time is further calculated based on the prediction covariance: ; In the formula, for The prior error covariance matrix at each time step is used to characterize the uncertainty of the current state prediction result before the fusion of new observations; The observation noise covariance matrix is ​​used to characterize the statistical level of vehicle longitudinal deceleration measurement noise and external disturbances. The Kalman gain at the current time step is used to determine the current observation residual. The weight of its role in the current cycle's state update. If A larger value indicates that the current observation information has a higher weight in this update cycle; if A smaller value indicates that the algorithm still largely maintains the old state. Among these, The parameters can be preset based on the static calibration results of the deceleration sensor, the statistical results of historical test data, or empirical calibration parameters, and in this embodiment, they are used as known filter parameters to participate in the Kalman gain solution and the subsequent residual determination quantity construction.

[0060] Further utilizing the Kalman gain and the observation residual at the current time, the prior stiffness estimate is updated to obtain the posterior estimate at the current time: ; In the formula, The posterior stiffness estimate (posterior estimate of the longitudinal slip stiffness of the road surface) after filtering correction at the current time is a characteristic quantity of the road surface adhesion condition, abbreviated as In standard Kalman filtering The solution is not given directly by the state equation all at once, but is completed by a recursive process of "prior prediction - residual calculation - gain solution - state update".

[0061] When road surface adhesion conditions remain stable, the standard Kalman filter can continuously estimate the characteristics of road surface adhesion conditions according to the above recursive process. However, when road surface adhesion conditions change abruptly, the error covariance matrix in the standard Kalman filter will gradually shrink in the steady-state phase, thus leading to a decrease in the Kalman gain. The Kalman gain decreases, reducing the weight of the current observation residual in the state update equation, thus affecting the posterior estimate. The response to new road surface conditions is slower, which is manifested as a lag in the updating of characterization parameters.

[0062] To address the aforementioned issues, this solution does not alter the state definition, observation definition, or final output format of the standard Kalman filter. Instead, it enhances the "prediction covariance update" step of the standard Kalman filter. First, it optimizes the observed residuals... Perform sliding window statistics to obtain the average covariance of the residual sequence. Then by Construction Decision Quantity and And based on this, the adaptive fading factor is obtained. The specific formula is as follows: ; ; ; .

[0063] In the formula, It is used to transform the residual at a single moment into a persistent mismatch statistic over a period of time, thereby distinguishing whether the current deviation belongs to a single random noise or has already reflected a change in the road surface adhesion conditions; This is the first criterion, used to characterize the extent to which the current residual statistic exceeds the observation noise baseline; The second criterion is used to characterize the magnitude of the projection of the prediction covariance into the observation space; It is an adaptive fading factor; This represents the trace operation of a matrix. In the one-dimensional observation system of this embodiment, , , and In numerical implementation, they all degenerate into scalars; therefore, the trace operation in the formula in this embodiment corresponds to the scalar itself. Since The observation noise covariance matrix is ​​directly introduced into the construction. ,therefore It also serves as the statistical baseline for observation noise, used to determine whether the current residual deviation has exceeded the interpretable range of normal measurement noise.

[0064] Obtaining the adaptive fading factor Then, it is used to correct the prediction covariance update process: ; In the formula, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. Let be the process noise covariance matrix. Where, This parameter is used to characterize the slow changes in road surface conditions and the statistical uncertainty caused by model approximation errors. It can be preset based on the system's dynamic response characteristics, historical estimation error statistics, or empirical calibration parameters. Due to the corrected... This will directly enter the Kalman gain. The solution to adaptively correct the predicted covariance essentially involves changing the Kalman gain to further alter... The update range.

[0065] Therefore, this invention enhances the covariance update stage of the standard Kalman filter. Its complete action chain is as follows: , When the road surface adhesion conditions change abruptly, Enlargement will lead to It is magnified, thus making Increase the weight of current observation information in state updates, thereby improving the final output. It can more quickly break free from the inertia of the old road surface and converge to the new road surface adhesion conditions.

[0066] The multidimensional feature confidence integration module is used to... As the core input for integral determination, and combined with , , and brake air pressure Auxiliary signals are used to filter, accumulate, and classify evidence for the current period. Used to characterize whether the road surface tends to have high, medium, or low adhesion in the current cycle. , , and It is mainly used to complete the integration system enable determination, abnormal operating condition suppression, and integration increment correction. The output of this step is the preliminary front axle adhesion category determination. Preliminary attachment category determination of the rear axle .

[0067] The calculation conditions for road surface adhesion condition characterization parameters include: Exceeding the set speed threshold to avoid low-speed measurement noise during vehicle start-up or braking. Brake air pressure. The pressure is above the set braking trigger threshold, and the pressure change gradient remains stable to eliminate false triggering interference. It is within the algorithm's reliable estimation range to ensure... It remains within the linear identification range that has a clear physical meaning.

[0068] An update to the score is only allowed if all of the above conditions are met; otherwise, the score from the previous period will remain unchanged.

[0069] The formula for calculating the initial discrimination score of road surface conditions is: In the formula, for The cumulative road surface condition score over time, i.e. the confidence score, is used to reflect whether the accumulated evidence up to the current time is more inclined to support the high adhesion, medium adhesion, or low adhesion category. The score correction value, i.e., the current period's score increment, is used to characterize the effect of newly added evidence in this period on strengthening, weakening, or maintaining the existing judgment trend.

[0070] As the primary criterion for updating the initial discrimination score of road surface conditions. When When the data remains within the first preset data range, it is determined that the current period's primary evidence supports a high-attachment direction, thus increasing the integral increment. Take the positive; when When the data remains within the third preset data range, the current period's primary evidence is determined to support the low-attachment direction, thus increasing the integral increment. Take the negative; when When the data is within the second preset data range, the attachment direction in the current period's primary evidence support is determined. Remain unchanged or make only minor adjustments.

[0071] In the After determining the principal direction of integration, then combine it with... and The matching relationship can enhance, weaken, or suppress the integral increment. Specifically, if Pointing towards the direction of high adhesion, and If the synchronization exhibits a relatively strong dynamic response, then positive scoring can be applied, denoted as... ;like Rapid growth If the increase is not synchronized, it indicates a risk of abnormal slip amplification or noise disturbance. In this case, a negative correction can be performed, denoted as... .

[0072] To reflect the principle of prioritizing braking safety, the following can be set: Greater than This allows the system to have a faster integral descent rate when signs of low adhesion risk appear, while maintaining a relatively smooth integral rise process when high adhesion is identified.

[0073] Set the first road surface condition adhesion determination threshold Second road surface condition adhesion determination threshold ,in By integrating and accumulating evidence from multiple consecutive periods, rather than directly determining based on instantaneous features at a single moment, the temporal continuity and noise resistance of the attachment category output can be improved.

[0074] when At that time, the "high adhesion" road surface category is initially determined.

[0075] when At that time, the output will preliminarily determine the "low adhesion" road surface category.

[0076] when At that time, a preliminary determination of the "medium adhesion" pavement category is output, and hysteresis holding can be introduced by combining the determination result of the previous time step. Thus, a preliminary front axle adhesion category determination is output. Preliminary attachment category determination of the rear axle It is not the stiffness estimate itself. If If the determination result is true, it can be directly used as the final road surface adhesion category identification result for the entire vehicle; if Then and Perform physical consistency correction.

[0077] The dynamic axle load collaborative correction module does not perform a separate road surface identification process again. Instead, when the initial adhesion category determination results of the front and rear axles are inconsistent, it performs a physical consistency check on the initial determination of the rear axle to identify the pseudo-slip phenomenon of the rear axle caused by brake load transfer. When the initial adhesion category determination results of the front and rear axles are consistent, it can... .

[0078] and Peak slip ratio of rear axle , The calculation includes structural parameters such as vehicle center of gravity height, wheelbase, vehicle mass, and static normal loads on the front and rear axles. During braking, the complete normal force model, including the increase in front axle load and the decrease in rear axle load, is calculated as follows: ; in, The front axle real-time normal load represents the actual load level on the front axle caused by longitudinal load transfer during braking. The front axle static normal load represents the reference load borne by the front axle when the vehicle is stationary or traveling in a straight line at a constant speed. For the overall vehicle weight; For longitudinal deceleration of the vehicle; The height of the vehicle's center of gravity; The vehicle's wheelbase is used for calculation. By calculating the dynamic load on the front axle, it can be combined with the dynamic load on the rear axle to describe the axle load redistribution during the vehicle's braking process, thus providing a physical basis for determining whether the difference in slip between the front and rear axles stems from differences in road surface adhesion conditions or from load transfer.

[0079] ; in, The real-time normal load on the rear axle represents the actual load level on the rear axle during braking. This is the static load on the rear axle, representing the load level on the rear axle under reference conditions; For the overall vehicle weight; For longitudinal deceleration of the vehicle; The height of the vehicle's center of gravity; This refers to the vehicle's wheelbase. The vehicle's center of gravity height, wheelbase, and overall vehicle mass collectively determine the magnitude of longitudinal load transfer under braking deceleration. If... relatively A significant decrease indicates that the rear axle is under reduced load at the current moment, making it more susceptible to transient high slippage under the same road conditions. Therefore, the dynamic load on the rear axle is a key physical quantity for determining whether pseudo-slippage exists.

[0080] When the initial adhesion classification of the front and rear axles is inconsistent, a collaborative verification is triggered. Priority should be given to resolving conflicting scenarios where the front axle is classified as having high adhesion while the rear axle is classified as having low adhesion, as this scenario best reflects the risk of rear axle pseudo-slip caused by brake load transfer. Record the peak rear axle slip rate. ,in This represents the peak slip ratio of the rear axle during the current braking process. If compared to If the decrease in axle load exceeds the set criterion, the rear axle is determined to be under light load. Therefore, the rear axle identification deviation is identified as a transient disturbance-induced pseudo-slip phenomenon caused by the reduced axle load. The initial low-adhesion determination of the rear axle is corrected to high-adhesion, resulting in a corrected rear axle adhesion category determination. And maintain the road surface category determination synchronized with the front axle. If the dynamic load reduction of the rear axle does not meet the correction conditions, then maintain... Unchanged. For situations where there are inconsistencies in other categories between the front and rear axles, the initial judgment result can be maintained, or the correction rules can be expanded according to the same physical consistency principle. The final road surface adhesion category identification result for the entire vehicle is formed by combining the front axle judgment result. .

[0081] The sliding window length in this scheme The adaptive factor weights and scoring thresholds can be fine-tuned according to the dynamic characteristics of different vehicle models.

[0082] Example 3 Figure 3 A schematic diagram of a vehicle-mounted device that can be used to implement embodiments of the present invention is shown. The components, their connections and relationships, and their functions shown herein are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0083] like Figure 3 As shown, the vehicle-mounted device 10 includes at least one processor 11 and a memory, such as ROM 12 or RAM 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from the storage unit 18. The RAM 13 can also store various programs and data required for the operation of the vehicle-mounted device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An I / O interface 15 is also connected to the bus 14. The ROM 12 is a read-only memory, the RAM 13 is a random access memory, and the I / O interface 15 is an input / output interface.

[0084] Multiple components in the vehicle-mounted device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, optical disk, etc.; and a communication unit 19, such as a network card, modem, wireless transceiver, etc. The communication unit 19 allows the vehicle-mounted device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0085] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as methods for identifying road surface adhesion conditions.

[0086] In some embodiments, the method for identifying road surface adhesion conditions may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the vehicle-mounted device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the method for identifying road surface adhesion conditions described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the method for identifying road surface adhesion conditions by any other suitable means (e.g., by means of firmware).

[0087] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0088] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0089] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable 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. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, 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.

[0090] To provide interaction with the user, the systems and techniques described herein can be implemented in an in-vehicle device having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the in-vehicle device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0091] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0092] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS servers, such as high management difficulty and weak business scalability.

[0093] This application also discloses a computer program product, which includes a computer program that, when executed by a processor, implements the method for identifying road surface adhesion conditions provided in any embodiment of this application. This program product and the methods for identifying road surface adhesion conditions disclosed in the embodiments of this application belong to the same inventive concept, and therefore will not be described in detail here.

[0094] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0095] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for identifying road surface adhesion conditions, characterized in that, include: Using the longitudinal slip stiffness of the road surface as the state variable and the longitudinal deceleration of the vehicle as the observation, the observation residuals are calculated using the Kalman filter algorithm, and the observation residuals are statistically analyzed using a sliding window to obtain the average covariance of the residual sequence. Based on the mean covariance of the residual sequence, the covariance matrix of the observation noise, the longitudinal slip rate of the wheel, and the covariance matrix of the prior error, two decision quantities are constructed, and the adaptive fading factor is determined based on the comparison results of the decision quantities. Based on the adaptive fading factor, the prior error covariance matrix is ​​corrected to obtain the corrected prior error covariance matrix; Based on the corrected prior error covariance matrix, observation residuals, and the current time-time prior stiffness estimate, calculate the road surface adhesion condition characterization quantity; Based on the road surface adhesion condition characterization quantity, the first road surface adhesion judgment threshold, and the second road surface adhesion judgment threshold, the front axle and rear axle adhesion category judgment results are determined, and the vehicle road surface adhesion conditions are determined based on the front axle and rear axle adhesion category judgment results.

2. The method according to claim 1, characterized in that, Before calculating the observation residuals using the Kalman filter algorithm with the road surface longitudinal slip stiffness as the state variable and the vehicle longitudinal deceleration as the observation, the following steps are also included: Based on the vehicle's yaw angle, the wheel speed is corrected to obtain the wheel speed compensation value; The longitudinal slip ratio of the wheels is calculated based on the wheel speed compensation value and the vehicle reference speed.

3. The method according to claim 1, characterized in that, Based on the mean covariance of the residual sequence, the covariance matrix of the observation noise, the longitudinal slip rate of the wheel, and the covariance matrix of the prior error, two decision quantities are constructed, including: The difference between the average covariance of the residual sequence and the covariance matrix of the observation noise is used as the first criterion. The second decision quantity is determined based on the wheel longitudinal slip ratio and the prior error covariance matrix.

4. The method according to claim 1, characterized in that, Based on the adaptive fading factor, the prior error covariance matrix is ​​corrected to obtain the corrected prior error covariance matrix, including: The prior error covariance matrix is ​​corrected based on the following formula: ； in, for The posterior error covariance matrix at time step [time]. Here is the state transition matrix. The process noise covariance matrix is... The adaptive fading factor is... Let be the corrected prior error covariance matrix.

5. The method according to claim 1, characterized in that, Based on the corrected prior error covariance matrix, observation residuals, and the current prior stiffness estimate, the road surface adhesion condition characterization quantities are calculated, including: Calculate the Kalman gain based on the corrected prior error covariance matrix, the wheel longitudinal slip ratio, and the observation noise covariance matrix. The product of the Kalman gain and the observation residual at the current time is summed with the prior stiffness estimate at the current time to obtain the road surface adhesion condition characterization quantity.

6. The method according to claim 1, characterized in that, Based on the road surface adhesion condition characterization quantity, the first road surface adhesion judgment threshold, and the second road surface adhesion judgment threshold, the front axle and rear axle adhesion category judgment results are determined, including: When the vehicle condition meets the calculation conditions for the road surface adhesion condition characterization quantity, the scoring correction value is determined according to the range of the road surface adhesion condition characterization quantity. The score correction value is summed with the cumulative road condition score of the previous time to obtain the initial road condition discrimination score; Based on the initial discrimination score of the road surface conditions, the first road surface condition adhesion judgment threshold, and the second road surface condition adhesion judgment threshold, the front axle and rear axle adhesion category judgment results are determined.

7. The method according to claim 1, characterized in that, Based on the front and rear axle adhesion category determination results, the overall vehicle road surface adhesion conditions are determined, including: If the front axle and rear axle attachment categories are inconsistent, and the front axle attachment category is the first attachment type while the rear axle attachment category is the second attachment type, if it is determined that there is pseudo-slippage in the rear axle dynamic load based on the real-time normal load and the static load of the rear axle, then the rear axle attachment category will be corrected to the first attachment type.

8. A vehicle-mounted device, characterized in that, The vehicle-mounted equipment includes: At least one processor, and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the method for identifying road surface adhesion conditions according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for identifying road surface adhesion conditions as described in any one of claims 1-7.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method for identifying road surface adhesion conditions according to any one of claims 1-7.

Images