Vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback

Abstract

The present application relates to the technical field of vehicle braking control, in particular to a vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback, comprising: processing the longitudinal acceleration signal through a recursive observer to obtain an accurate vehicle mass center estimated longitudinal deceleration, and dynamically correcting the real-time slip rate calculated based on the wheel speed with the same to obtain a corrected slip rate that more accurately reflects the current road adhesion state. At the same time, the real-time wheel speed difference ratio of the coaxial wheels is calculated, which is fused with the corrected slip rate in a decision model to jointly decide the target braking pressure of each wheel, and finally to realize independent closed-loop adjustment of the braking pressure of each wheel. The method can effectively improve the identification accuracy and adaptability of the system to different adhesion road surfaces, and take into account the braking efficiency and directional stability during braking.

Description

Technical Field

[0001] This invention relates to the field of vehicle braking control technology, and in particular to a vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback. Background Technology

[0002] Existing anti-lock braking systems (ABS) for vehicles generally employ a control strategy that calculates the slip ratio based on wheel speed and a reference vehicle speed, and adjusts pressure by comparing logical threshold values. The core of this approach lies in determining the wheel motion state, and the accuracy of this determination heavily depends on the precision of the reference vehicle speed and slip ratio calculations. During actual braking, due to the influence of vehicle pitch and sensor mounting positions, the signals collected by the longitudinal acceleration sensors contain significant noise and interference, failing to directly and accurately reflect the true motion state of the vehicle's center of gravity. Directly using this raw deceleration signal or employing a fixed threshold can lead to lag or even misjudgment in the system's identification of the road surface adhesion coefficient. Under complex conditions such as low adhesion or split-slip surfaces, the control robustness is insufficient, making it difficult to achieve optimal slip ratio tracking.

[0003] Conventional anti-lock braking systems (ABS) in vehicles often rely on the slip ratio of individual wheels or apply largely consistent control logic to wheels on the same axle when determining braking pressure. However, in real-world road conditions, especially on curves or surfaces with asymmetrical coefficients of friction, the adhesion conditions of the left and right wheels on the same axle differ. Relying solely on single-wheel slip ratio information cannot effectively identify and address these differences, potentially leading to irrational distribution of braking force on both sides, affecting the vehicle's braking directional stability and steering control. Current technology lacks an effective mechanism that can integrate the overall vehicle deceleration state with information on the motion differences between wheels on the same axle for joint and refined pressure decision-making. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and propose a vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback, comprising: The real-time rotational speed of each wheel of the vehicle is obtained, and the instantaneous wheel speed of each wheel is calculated. Based on the instantaneous wheel speed of each wheel, the overall reference speed of the vehicle is calculated in real time. The instantaneous wheel speed of each wheel is compared with the reference vehicle speed to calculate the real-time slip ratio of the corresponding wheel; Simultaneously acquire actual longitudinal deceleration signals from the vehicle's longitudinal acceleration sensor; The actual longitudinal deceleration signal is input into the recursive observer to calculate the estimated longitudinal deceleration of the vehicle's center of gravity. Based on the real-time slip ratio of each wheel, the wheel speed difference between the left and right wheels on the same axle is calculated, and the real-time wheel speed difference ratio between the wheels on the same axle is further calculated. The estimated longitudinal deceleration, the real-time slip ratio of each wheel, and the real-time wheel speed difference ratio between wheels on the same axle are all input into the braking pressure decision model. Within the braking pressure decision model, the real-time slip ratio is dynamically corrected based on the estimated longitudinal deceleration to obtain a corrected slip ratio adapted to the current road conditions. Within the braking pressure decision model, the corrected slip ratio and the real-time wheel speed difference ratio are combined to jointly determine the independent target braking pressure value for each wheel. Based on the target braking pressure value for each wheel, an independent braking pressure adjustment command is generated to control the corresponding solenoid valve to adjust the actual pressure of each wheel's brake caliper.

[0006] As a further aspect of the present invention, the step of calculating the overall reference vehicle speed in real time based on the instantaneous wheel speeds of each wheel includes: The maximum value among the instantaneous wheel speeds of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel is selected as the initial reference wheel speed. The reference wheel acceleration is obtained by performing a time-based differential operation on the initial reference wheel speed; The acceleration of the reference wheel is compared with a preset acceleration threshold. When the acceleration of the reference wheel is greater than the acceleration threshold, it is determined that the reference wheel is in a state of violent deceleration. During the period when the reference wheel is in a state of severe deceleration, the effective reference vehicle speed calculated in the previous control cycle is used, and the longitudinal deceleration of the vehicle in the previous control cycle is combined to extrapolate and calculate the reference vehicle speed for the current control cycle. When the reference wheel is not in a state of drastic deceleration, the initial reference wheel speed is directly used as the reference vehicle speed for the current control cycle.

[0007] As a further aspect of the present invention, the actual longitudinal deceleration signal is input into a recursive observer to calculate the estimated longitudinal deceleration of the vehicle's center of gravity, including: Establish a simplified longitudinal dynamics model of the vehicle that includes vehicle mass, center of gravity height, wheelbase, and wheel rolling radius; In the recursive observer, the actual longitudinal deceleration signal is used as the input excitation; Based on the simplified vehicle longitudinal dynamics model, state equations and observation equations are constructed in the recursive observer, where the state variables include the longitudinal velocity and longitudinal deceleration of the vehicle's center of mass. Using the algorithm of the recursive observer, the longitudinal velocity and longitudinal deceleration of the vehicle's center of gravity in the current cycle are recursively calculated by combining the state estimate of the previous cycle with the actual longitudinal deceleration signal input in the current cycle. The output longitudinal deceleration is the estimated longitudinal deceleration.

[0008] As a further aspect of the present invention, the step of calculating the wheel speed difference between the left and right wheels on the same axle based on the real-time slip ratio of each wheel, and further calculating the real-time wheel speed difference ratio between the wheels on the same axle, includes: For the same axle of the vehicle, read the instantaneous wheel speed of the left wheel and the instantaneous wheel speed of the right wheel; Calculate the absolute value of the difference between the instantaneous wheel speed of the left wheel and the instantaneous wheel speed of the right wheel to obtain the original value of the wheel speed difference of the wheels on the same axle; Read the instantaneous wheel speed of the wheel with the higher wheel speed in the coaxial section as the wheel speed reference; Divide the original value of the wheel speed difference between the coaxial wheels by the wheel speed reference to obtain the normalized wheel speed difference ratio, which is the real-time wheel speed difference ratio between the coaxial wheels.

[0009] As a further aspect of the present invention, within the braking pressure decision model, the real-time slip ratio is dynamically corrected based on the estimated longitudinal deceleration to obtain a corrected slip ratio adapted to the current road conditions, including: Within the braking pressure decision model, an ideal slip ratio range under standard road conditions is preset; The estimated longitudinal deceleration is matched with the pre-stored typical deceleration feature maps under different road surface adhesion coefficients to infer the equivalent adhesion level of the current road surface. Based on the inferred equivalent adhesion level, the upper and lower boundaries of the ideal slip ratio range are dynamically adjusted to generate the target slip ratio band adapted to the current road surface. The slip ratio of the wheel is compared with the center value of the target slip ratio band to calculate the slip ratio deviation; Based on the slip ratio deviation, a correction function related to the estimated longitudinal deceleration is applied to the real-time slip ratio, and the corrected slip ratio is output. The steps for constructing the braking pressure decision model include: Collect actual braking process data of vehicles under different road surface adhesion coefficients and different braking intensities. The actual braking process data includes wheel speed signals, vehicle longitudinal deceleration signals and corresponding brake caliper pressure values. Feature extraction is performed on the collected actual braking process data to obtain the wheel speed change curve, slip ratio change curve, and coaxial wheel speed difference ratio curve for each working condition; The estimated longitudinal deceleration of the vehicle under each working condition is used as the input variable, and the slip ratio range corresponding to maintaining the wheel in the stable region during braking is used as the output label to form the first training dataset. The wheel speed difference ratio under each working condition is used as the input variable, and the direction and magnitude of the left and right wheel pressure difference required to maintain vehicle stability are used as the output labels to form the second training dataset. Based on the first training dataset, a slip ratio dynamic correction sub-network model is trained; Based on the second training dataset, a wheel speed difference balance sub-network model is trained; The slip ratio dynamic correction subnetwork model and the wheel speed difference balance subnetwork model are connected in parallel, and the output layers of the two subnetworks are merged into a common pressure decision fully connected layer to form the complete braking pressure decision model. The trained braking pressure decision model is then stored in the storage medium of the vehicle's electronic control unit.

[0010] As a further aspect of the present invention, based on the slip ratio deviation, a correction function related to the estimated longitudinal deceleration is applied to the real-time slip ratio to output the corrected slip ratio, including: Define a slip ratio correction factor, which is a monotonically decreasing function of the estimated longitudinal deceleration; When the estimated longitudinal deceleration is large, it indicates that the road surface has good adhesion, the slip ratio correction coefficient is small, and the correction range of the real-time slip ratio is small. When the estimated longitudinal deceleration is small, it indicates poor road adhesion. A large value for the slip ratio correction coefficient results in a large correction range for the real-time slip ratio. Multiply the slip ratio deviation by the slip ratio correction coefficient to obtain the correction amount, and add the correction amount to the real-time slip ratio to obtain the corrected slip ratio.

[0011] As a further aspect of the present invention, within the braking pressure decision model, the corrected slip ratio and the real-time wheel speed difference ratio are combined to jointly determine an independent target braking pressure value for each wheel, including: A base pressure value is set for each wheel, and the base pressure value is determined by the driver's brake pedal travel. A pressure adjustment factor is preset for each wheel, and the initial value of the pressure adjustment factor is zero; The corrected slip ratio of the wheel is compared with a preset slip ratio threshold, and the pressure adjustment factor is updated based on the comparison result. The real-time wheel speed difference ratio between coaxial wheels is compared with a preset wheel speed difference threshold, and the pressure adjustment factor is updated based on the comparison result. The pressure adjustment factor, which has been updated twice, is weighted and synthesized with the base pressure value to generate the target braking pressure value of the wheel.

[0012] As a further aspect of the present invention, the step of comparing the corrected slip ratio of the wheel with a preset slip ratio threshold and updating the pressure adjustment factor based on the comparison result includes: Preset high and low slip ratio thresholds for each wheel; When the corrected slip ratio of the wheel is greater than the slip ratio high threshold, the wheel is determined to be slipping excessively, and a negative adjustment amount is superimposed on the pressure adjustment factor; When the corrected slip ratio of the wheel is less than the low slip ratio threshold, the wheel is determined to be slipping insufficiently, and a positive adjustment amount is added to the pressure adjustment factor. When the wheel's corrected slip ratio is between the low slip ratio threshold and the high slip ratio threshold, the pressure adjustment factor is not updated based on the slip ratio.

[0013] As a further aspect of the present invention, the step of comparing the real-time wheel speed difference ratio between coaxial wheels with a preset wheel speed difference threshold, and updating the pressure adjustment factor based on the comparison result, includes: A threshold value for the wheel speed difference ratio is preset for the left and right wheels on the same axle; Compare the real-time wheel speed difference ratio of the left and right wheels on the same axle with the wheel speed difference ratio threshold; When the real-time wheel speed difference ratio is greater than the wheel speed difference ratio threshold, it is determined that the difference in adhesion between the wheels on both sides of the axle is significant. A positive adjustment amount is added to the pressure adjustment factor of the wheel on the side with lower wheel speed, and a negative adjustment amount is added to the pressure adjustment factor of the wheel on the side with higher wheel speed, or the pressure adjustment factor is kept unchanged. When the real-time wheel speed difference ratio is less than or equal to the wheel speed difference ratio threshold, the pressure adjustment factor of any wheel on the axle is not updated based on the wheel speed difference ratio.

[0014] As a further aspect of the present invention, the step of generating an independent brake pressure adjustment command based on the target brake pressure value of each wheel, and controlling the corresponding solenoid valve to adjust the actual pressure of each wheel brake caliper, includes: Real-time acquisition of the actual pressure value of each wheel brake caliper; The actual pressure value of each wheel is compared with the target braking pressure value of the corresponding wheel to calculate the pressure deviation value; Based on the sign and magnitude of the pressure deviation value, the working state of the corresponding solenoid valve is determined, including pressure boosting state, pressure holding state, and pressure reducing state. The determined solenoid valve operating state is encoded into a specific pulse width modulation signal, which is then sent to the solenoid valve driver of the corresponding wheel as the brake pressure adjustment command. The solenoid valve is controlled by the solenoid valve driver to open and close the solenoid valve, thereby adjusting the brake fluid flow rate so that the actual pressure value approaches the target brake pressure value.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: By processing the actual longitudinal deceleration signal from the sensors into the recursive observer, high-frequency noise and interference components caused by changes in vehicle attitude can be effectively filtered out, resulting in an estimated longitudinal deceleration that more closely approximates the true motion of the vehicle's center of gravity. This accurate estimate is then used to dynamically correct the real-time slip ratio calculated based on wheel speed, with the correction process responding in real-time to changes in the overall vehicle deceleration. This makes the corrected slip ratio more accurately reflect the actual adhesion state between the tires and the road surface, reducing control errors caused by inaccurate sensor signals or vehicle pitch. The control system responds based on more accurate tire-road condition information, improving the real-time performance and accuracy of its decisions. Especially during unsteady braking processes with drastic deceleration changes, the system can identify road adhesion trends earlier and more stably.

[0016] After correcting the slip ratio based on the estimated longitudinal deceleration, the decision model incorporates the real-time wheel speed difference ratio between the wheels on the same axle for joint decision-making. The wheel speed difference ratio directly quantifies the degree of speed difference between the wheels on both sides of the same axle due to different adhesion conditions. By fusing the corrected slip ratio, which reflects the overall braking intensity, with the wheel speed difference ratio, which reflects the difference in adhesion between the two sides, the model can set differentiated target braking pressures for the left and right wheels on the same axle. This decision-making mechanism enables the system not only to prevent individual wheel lock-up but also to actively adjust the braking force on both sides to balance the vehicle's yaw motion. When braking on high- or low-friction surfaces, the system prioritizes braking efficiency; when braking on split surfaces or curves, the system automatically adjusts the pressure distribution based on the wheel speed difference ratio to suppress the vehicle's tendency to veer or fishtail, thereby maintaining a shorter braking distance while enhancing directional stability and the driver's steering control during braking. Attached Figure Description

[0017] Figure 1 This is a flowchart of the vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback as described in this invention; Figure 2 A flowchart for dynamic calculation of reference vehicle speed; Figure 3A flowchart for calculating the wheel speed difference ratio of coaxial wheels; Figure 4 This refers to the change in the ratio of longitudinal deceleration to wheel speed difference during braking. Figure 5 Comparison curves before and after correction of the left front wheel slip ratio. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] See Figure 1 The system acquires the real-time rotational speed of each wheel of the vehicle and calculates the instantaneous wheel speed accordingly. Based on these instantaneous wheel speeds, it calculates a reference speed representing the overall motion state of the vehicle. By comparing the instantaneous wheel speed of each wheel with the reference speed, the real-time slip ratio of the corresponding wheel can be calculated. Simultaneously with acquiring the wheel motion state, the method also collects actual longitudinal deceleration signals from the vehicle's longitudinal acceleration sensor. This actual longitudinal deceleration signal is input into a recursive observer, and after calculation, an estimated longitudinal deceleration that more accurately reflects the motion state of the vehicle's center of gravity is obtained. On the other hand, based on the real-time slip ratio of each wheel, the method calculates the wheel speed difference between the left and right wheels on the same axle, and further calculates the normalized ratio of the real-time wheel speed difference between the wheels on the same axle. The estimated longitudinal deceleration, the real-time slip ratio of each wheel, and the ratio of the real-time wheel speed difference between the wheels on the same axle are input into a pre-built braking pressure decision model. Within this model, the real-time slip ratio is first dynamically corrected based on the estimated longitudinal deceleration to obtain a corrected slip ratio adapted to the current road surface adhesion conditions. The model combines the corrected slip ratio with the real-time wheel speed difference ratio to make joint decisions and outputs an independent target braking pressure value for each wheel. Based on the target braking pressure value for each wheel, the system generates an independent braking pressure adjustment command, controls the corresponding solenoid valve to adjust the actual pressure of each wheel's brake caliper, and achieves anti-lock braking control.

[0021] See Figure 2 In one embodiment of the present invention, a scenario is considered where a vehicle is undergoing emergency braking under two different adhesion conditions: dry asphalt road surface and compacted snow road surface. In the dry asphalt road surface scenario, the instantaneous wheel speeds of the four wheels of the vehicle at the initial stage of braking are: 18.5 revolutions per second for the left front wheel, 18.3 revolutions per second for the right front wheel, 19.1 revolutions per second for the left rear wheel, and 18.9 revolutions per second for the right rear wheel. The control system selects the maximum value among these four instantaneous wheel speeds, namely 19.1 revolutions per second for the left rear wheel, as the initial reference wheel speed. A time-based differential operation is performed on the initial reference wheel speed. Assuming that the initial reference wheel speed decreases to 18.9 revolutions per second within a millisecond control cycle, the calculated reference wheel acceleration is -200 revolutions per second. The preset acceleration threshold is -250 revolutions per second. Since the absolute value of the reference wheel acceleration of -200 revolutions per second is less than the absolute value of the preset acceleration threshold of -250 revolutions per second, it is determined that the reference wheel is not in a state of violent deceleration. Therefore, the reference vehicle speed in the current control cycle is directly assigned the initial reference wheel speed of 19.1 revolutions per second.

[0022] In some embodiments, when a vehicle is traveling on compacted snow, the wheels are more prone to locking up due to the low coefficient of friction. At the initial moment of braking, the instantaneous wheel speeds may be: 16.0 revolutions per second for the left front wheel, 15.8 revolutions per second for the right front wheel, 17.5 revolutions per second for the left rear wheel, and 16.5 revolutions per second for the right rear wheel. The maximum value, 17.5 revolutions per second for the left rear wheel, is selected as the initial reference wheel speed. In the next millisecond control cycle, the instantaneous wheel speed of the left rear wheel may drop sharply to 14.0 revolutions per second, resulting in a calculated reference wheel acceleration of -3500 revolutions per second. Since the absolute value of this reference wheel acceleration is much greater than the preset acceleration threshold of -250 revolutions per second, it is determined that the left rear wheel, serving as the reference, is undergoing severe deceleration. During periods of rapid deceleration, the calculation of the reference vehicle speed no longer relies on the potentially distorted maximum instantaneous wheel speed. Instead, it uses the effective reference vehicle speed calculated and saved from the previous control cycle. Assuming the effective reference vehicle speed of the previous control cycle is 18.0 revolutions per second, and the longitudinal deceleration of the vehicle in the previous control cycle is also obtained, assumed to be negative 6.5 meters per second, the reference vehicle speed for the current control cycle is calculated using the following extrapolation formula: ; in: This indicates the reference vehicle speed for the current control cycle. This indicates the effective reference speed of the vehicle in the previous control cycle. This represents the longitudinal deceleration in the previous control cycle. This indicates the duration of the control cycle. Substitute the value... Rotational speed per second. This method can be understood as preventing the reference vehicle speed signal from dropping abnormally when the reference wheel experiences a sudden drop in wheel speed due to abrupt changes in adhesion or impact from uneven road surfaces, thus providing a relatively smooth and reasonable benchmark for calculating the slip ratio of other wheels.

[0023] Optionally, the preset acceleration threshold is a fixed value calibrated based on vehicle dynamics characteristics. In another embodiment, the preset acceleration threshold can also be designed as a dynamic variable associated with the initial reference wheel speed or the estimated longitudinal deceleration. For example, the preset acceleration threshold can be set as a function of the initial reference wheel speed. When the initial reference wheel speed is high, the absolute value of the preset acceleration threshold increases accordingly to accommodate a larger deceleration range that may occur during high-speed braking; when the initial reference wheel speed is low, the absolute value of the preset acceleration threshold decreases accordingly. In some embodiments, the preset acceleration threshold can also be linked to the estimated longitudinal deceleration, so that the criteria for determining a severe deceleration state can be adaptively adjusted according to the overall deceleration trend of the vehicle.

[0024] See Figure 3 In one embodiment of the present invention, considering a scenario where a vehicle brakes in a curve, the actual longitudinal deceleration signal from the vehicle's longitudinal acceleration sensor includes not only the deceleration caused by braking but also measurement interference caused by changes in vehicle attitude. To obtain a more accurate estimate of the vehicle's longitudinal deceleration from its center of gravity, a simplified vehicle longitudinal dynamics model needs to be established. This model includes parameters such as vehicle mass, center of gravity height, wheelbase, and wheel rolling radius. In the recursive observer, the actual longitudinal deceleration signal collected by the longitudinal acceleration sensor is used as the input excitation. Based on the simplified vehicle longitudinal dynamics model, state equations and observation equations are constructed within the recursive observer, where the state variables include the longitudinal velocity and longitudinal deceleration of the vehicle's center of gravity.

[0025] The algorithm of a recursive observer is used for calculation. The recursive observer combines the state estimate from the previous control cycle with the actual longitudinal deceleration signal input in the current cycle to recursively calculate the longitudinal velocity and longitudinal deceleration of the vehicle's center of gravity in the current cycle. The recursive calculation process can be represented by the following state update formula: ; in: This represents the state estimation vector for the kth control cycle, which includes the longitudinal velocity and longitudinal deceleration of the vehicle's center of gravity. This represents the state transition matrix from the (k-1)th control cycle to the kth control cycle, and its parameters are derived from the simplified vehicle longitudinal dynamics model. This represents the state estimation vector for the (k-1)th control cycle; This represents the observer gain matrix for the k-th control cycle; This represents the observed input for the k-th control cycle, i.e., the actual longitudinal deceleration signal; This represents the observation matrix. The longitudinal deceleration state variable calculated using the formula is the estimated longitudinal deceleration. It can be understood that the recursive observer can filter out high-frequency noise and interference introduced by vehicle pitch motion from the actual longitudinal deceleration signal based on the model, outputting an estimated longitudinal deceleration that better reflects the true motion state of the vehicle's center of gravity.

[0026] In some embodiments, a method for calculating the real-time wheel speed difference ratio between coaxial wheels is used to identify the adhesion difference between the left and right wheels on the same axle. For the front axle of a vehicle, the instantaneous wheel speeds of the left and right wheels are read. Assuming that during braking, due to the left wheel passing over a locally slippery surface, the instantaneous wheel speed of the left wheel is 12.5 revolutions per second, and the instantaneous wheel speed of the right wheel is 15.0 revolutions per second. The absolute value of the difference between the instantaneous wheel speeds of the left and right wheels is calculated, yielding the original value of the wheel speed difference between the coaxial wheels as |12.5 - 15.0| = 2.5 revolutions per second. The instantaneous wheel speed of the wheel with the higher speed on the coaxial wheel is read as the wheel speed reference, i.e., 15.0 revolutions per second for the right wheel. The original value of the wheel speed difference between the coaxial wheels is divided by the wheel speed reference, yielding a normalized wheel speed difference ratio of 2.5 / 15.0 ≈ 0.1667. This ratio is the real-time wheel speed difference ratio between the front axle wheels.

[0027] Optionally, in the calculation of the vehicle's rear axle, if the left and right wheels have uniform adhesion, the instantaneous wheel speed of the left wheel is 14.8 revolutions per second, and the instantaneous wheel speed of the right wheel is 14.9 revolutions per second. The calculated original value of the wheel speed difference is |14.8-14.9|=0.1 revolutions per second. The wheel speed benchmark is taken as 14.9 revolutions per second for the right wheel. The final calculated real-time wheel speed difference ratio is 0.1 / 14.9≈0.0067. It can be understood that by normalizing the wheel speed difference by dividing it by the higher wheel speed on the same axle, the real-time wheel speed difference ratio becomes a dimensionless relative value. Its magnitude directly reflects the degree of difference in the slip state of the wheels on both sides of the same axle, eliminating the influence of the different overall wheel speeds on the absolute value of the difference, allowing the subsequent decision model to make judgments based on a consistent scale.

[0028] In one embodiment of the present invention, the construction of the braking pressure decision model begins with data acquisition. Actual braking process data of the vehicle under different road surface adhesion coefficients and different braking intensities is collected. This actual braking process data includes wheel speed signals, vehicle longitudinal deceleration signals, and corresponding brake caliper pressure values. Taking emergency braking conditions on dry asphalt pavement (high adhesion) and compacted snow pavement (low adhesion) as examples, actual braking process data from the start of braking to the vehicle's stop is collected under both conditions. Feature extraction is performed on the collected actual braking process data to obtain the wheel speed variation curve, slip ratio variation curve, and co-axle wheel speed difference ratio curve for each condition.

[0029] The first training dataset is constructed by using the estimated longitudinal deceleration output by the recursive observer for each driving condition as input variables and the slip ratio range corresponding to maintaining the wheels in a stable region and achieving high braking efficiency during braking as output labels. For example, under the condition of dry asphalt pavement, the estimated longitudinal deceleration is relatively large, and its corresponding optimal slip ratio range may be 8% to 15%; under the condition of compacted snow pavement, the estimated longitudinal deceleration is relatively small, and its corresponding optimal slip ratio range may be 12% to 20%. Simultaneously, the real-time wheel speed difference ratio calculated for each driving condition is used as input variables, and the direction and magnitude of the left and right wheel pressure difference required to maintain a stable vehicle trajectory are used as output labels to construct the second training dataset. Based on the first training dataset, a slip ratio dynamic correction sub-network model is trained, which learns the mapping relationship from the estimated longitudinal deceleration to the target slip ratio range. Based on the second training dataset, a wheel speed difference balance sub-network model is trained, which learns the mapping relationship from the real-time wheel speed difference ratio to the left and right wheel braking force adjustment.

[0030] In some embodiments, the trained slip ratio dynamic correction subnetwork model and the wheel speed difference balance subnetwork model are connected in parallel, and the output layers of these two subnetworks are merged into a common pressure decision fully connected layer to form a complete braking pressure decision model. The pressure decision fully connected layer receives the features output by the two subnetworks and comprehensively determines the final target braking pressure value for each wheel. The trained braking pressure decision model is stored in the storage medium of the vehicle electronic control unit and directly invoked during real-time control. When performing dynamic slip ratio correction within the braking pressure decision model, an ideal slip ratio range under standard road conditions is preset within the model. The estimated longitudinal deceleration input in real time is matched with pre-stored typical deceleration feature maps under different road adhesion coefficients to infer the equivalent adhesion level of the current road surface.

[0031] Based on the inferred equivalent adhesion level, the upper and lower boundaries of the ideal slip ratio range are dynamically adjusted to generate a target slip ratio band suitable for the current road surface. For example, assuming the center value of the current target slip ratio band is 14%, and the real-time slip ratio of a certain wheel is 18%, the slip ratio deviation is +4%. Based on the slip ratio deviation, a correction function related to the estimated longitudinal deceleration is applied to the real-time slip ratio, outputting the corrected slip ratio. The correction function involves defining a slip ratio correction coefficient, which is a monotonically decreasing function of the estimated longitudinal deceleration. When the estimated longitudinal deceleration is large, it indicates good road surface adhesion, and the slip ratio correction coefficient is small, resulting in a small correction to the real-time slip ratio; when the estimated longitudinal deceleration is small, it indicates poor road surface adhesion, and the slip ratio correction coefficient is large, resulting in a large correction to the real-time slip ratio.

[0032] It is understandable that through the above mapping and correction process, the system can automatically adapt to changes in road surface. For example, when driving from an asphalt road onto ice, the estimated longitudinal deceleration will decrease, and the slip ratio correction coefficient will increase accordingly. The system will then more actively correct the slip ratio, thereby adopting a higher target slip ratio on low-adhesion surfaces to avoid wheel lock-up. Multiplying the slip ratio deviation by the slip ratio correction coefficient yields the correction amount. Adding the correction amount to the real-time slip ratio gives the corrected slip ratio. This relationship can be expressed by the following formula: ; in: Indicates the corrected slip ratio. Indicates the real-time slip ratio. Indicates the center value of the target slip ratio band. This indicates an estimate of the longitudinal deceleration. Indicates the estimation of longitudinal deceleration This is the slip ratio correction coefficient function for the independent variable, and this function is monotonically decreasing. Optional, the function... The specific form can be designed as a piecewise linear function or a smooth exponential decay function, and calibrated using actual vehicle test data. In some embodiments, the width of the target slip ratio band can also be dynamically adjusted based on the estimated longitudinal deceleration, using a narrower bandwidth in the high deceleration range to pursue precise control, and a wider bandwidth in the low deceleration range to enhance system stability. It can be understood that the braking pressure decision model and its internal dynamic slip ratio correction mechanism together achieve online identification and adaptive control of time-varying road conditions.

[0033] In one embodiment of the invention, the scenario of a vehicle's single wheel passing over a water film on a dry asphalt road surface is compared with the scenario of braking on a compacted snow road surface. A base pressure value is set for each wheel, which is determined by the driver's brake pedal travel. It is assumed that in the example braking scenario, the base pressure value corresponding to the driver pressing the brake pedal is 10 MPa. Simultaneously, a pressure adjustment factor is preset for each wheel, with an initial value of zero. The decision-making process first updates the pressure adjustment factor based on the corrected slip ratio. Each wheel has a preset high slip ratio threshold and a low slip ratio threshold. It is assumed that the high slip ratio threshold is set to 20%, and the low slip ratio threshold to 10%. In the scenario of a single wheel passing over a water film on a dry asphalt road surface, the vehicle's left front wheel may experience a decrease in adhesion due to the instantaneous passage over the water film, and its corrected slip ratio may reach 25%. This value is greater than the high slip ratio threshold of 20%, therefore, the left front wheel is determined to have excessive slip, and a negative adjustment amount is superimposed on the pressure adjustment factor of the left front wheel. It is assumed that the negative adjustment amount is -0.3. At the same time, the right front wheel is traveling on a dry road surface, and its corrected slip ratio is 12%. This value is between the low slip ratio threshold of 10% and the high slip ratio threshold of 20%, so the pressure adjustment factor of the right front wheel is not updated based on the slip ratio. For the coaxial rear wheels, their corrected slip ratios are all 8%, which is less than the low slip ratio threshold of 10%. It is determined that the rear wheels have insufficient slip, so a positive adjustment amount is added to the pressure adjustment factors of both the left and right rear wheels. Assuming the positive adjustment amount is +0.2, refer to Table 1.

[0034] Table 1. Pressure Adjustment Factor Update Process under Different Braking Scenarios: In some embodiments, the decision-making process then updates the pressure adjustment factor based on the real-time wheel speed difference ratio, and presets a wheel speed difference ratio threshold for the left and right wheels on the same axle, assuming the wheel speed difference ratio threshold is 0.1. The real-time wheel speed difference ratio of the left and right wheels on the same axle is compared with the wheel speed difference ratio threshold. In a scenario where water flows across one side of a dry asphalt road, due to a sudden drop in adhesion on the left front wheel, the real-time wheel speed difference ratio of the left and right wheels on the front axle may be 0.18. This value is greater than the wheel speed difference ratio threshold of 0.1, indicating a significant difference in adhesion between the two wheels on the front axle. At this time, a positive adjustment is added to the pressure adjustment factor of the wheel with the lower wheel speed (left front wheel) to attempt to increase its braking force and reduce slip, assuming the positive adjustment is +0.25; a negative adjustment is added to the pressure adjustment factor of the wheel with the higher wheel speed (right front wheel) to moderately reduce its braking force and balance the slip ratio on both sides, assuming the negative adjustment is -0.15. For the rear axle in this scenario, the real-time wheel speed difference ratio between its left and right wheels may be 0.01. This value is less than the wheel speed difference ratio threshold of 0.1, so the pressure adjustment factor of any wheel on the rear axle is not updated based on the wheel speed difference ratio.

[0035] Optionally, in a uniform braking scenario on a compacted snow surface, all wheels are in a low-adhesion state, and the corrected slip ratios of the left and right wheels on the same axle may be high and similar. For example, the corrected slip ratio of the left front wheel is 22%, and the corrected slip ratio of the right front wheel is 21%, both exceeding the high slip ratio threshold of 20%. Therefore, a negative adjustment of -0.3 is superimposed on the pressure adjustment factor of both front wheels. At this time, the real-time wheel speed difference ratio of the front axle may be 0.05, which is less than the wheel speed difference ratio threshold of 0.1. Therefore, no additional update is made to the pressure adjustment factor of the front wheels based on the wheel speed difference ratio. It can be understood that by independently updating the pressure adjustment factor based on the corrected slip ratio and the real-time wheel speed difference ratio, the system can simultaneously handle the requirements of longitudinal wheel slip control and lateral vehicle stability control. After completing two updates based on the corrected slip ratio and the real-time wheel speed difference ratio, the final pressure adjustment factor is weighted and synthesized with the base pressure value to generate an independent target braking pressure value for each wheel. The weighted synthesis can be achieved using the following formula: ; in: This represents the final target braking pressure value for the wheels; This indicates the base pressure value determined by the driver's brake pedal travel; This represents the pressure adjustment factor component after being updated based on the slip ratio; This represents the pressure adjustment factor component after being updated based on the wheel speed difference ratio; and These are the slip ratio adjustment weighting coefficient and the wheel speed difference adjustment weighting coefficient, used to adjust the relative influence of the two adjustment amounts. It can be understood that the weighting coefficients... and The system can be dynamically adjusted based on vehicle conditions. For example, a higher weight can be assigned to slip ratio adjustment during high-speed straight-line braking, while the weight of wheel speed difference adjustment can be increased during cornering or low-traction braking to enhance stability. In some embodiments, the base pressure value... It may not come directly from the pedal travel, but rather from a base pressure request value processed by the upper controller of the anti-lock braking system.

[0036] See Figure 4The figure presents the system response characteristics under different road surface adhesion conditions. From the curve characteristics: the longitudinal deceleration on dry roads (red triangle curve) rises rapidly in the initial stage of braking, stabilizing at approximately 9 m / s² after about 1.5 seconds, reflecting the high braking efficiency under high-adhesion road surfaces. The longitudinal deceleration on snowy roads (blue dot curve) increases gradually, eventually stabilizing at approximately 4.5 m / s², reflecting the braking capacity limitation under low-adhesion road surfaces. The real-time wheel speed difference ratio of the front axle (green square curve) continuously increases with braking time, approaching 0.07 at the end of braking, and remains consistently below the preset wheel speed difference ratio threshold (red dashed line, 0.10), indicating that the system effectively controls the adhesion difference between coaxial wheels through pressure regulation, maintaining the lateral stability of the vehicle body. The core value of the figure lies in its quantification of the impact of different road surface adhesion coefficients on braking performance, while verifying the effectiveness of the control method in balancing longitudinal braking efficiency and lateral stability: on dry roads, the system prioritizes high deceleration; on snowy roads, the system ensures stable and controllable braking by limiting deceleration and controlling wheel speed difference.

[0037] In one embodiment of the present invention, assuming that the target braking pressure value calculated by the braking pressure decision model for the left front wheel is 8.5 MPa, while the actual pressure value of the left front wheel brake caliper collected by the pressure sensor is 9.2 MPa. The actual pressure value of the left front wheel is compared with the target braking pressure value to calculate the pressure deviation value. The pressure deviation value is equal to the actual pressure value minus the target braking pressure value, i.e., 9.2 MPa minus 8.5 MPa, resulting in a pressure deviation value of +0.7 MPa. Based on the sign and magnitude of the pressure deviation value, the operating state of the corresponding solenoid valve is determined. A pressure deviation threshold is preset, for example, 0.5 MPa. When the calculated pressure deviation value is positive and its absolute value is greater than the pressure deviation threshold, it indicates that the actual pressure value is higher than the target braking pressure value, and the pressure needs to be reduced; the solenoid valve is then set to enter a depressurization state. When the pressure deviation value is negative and its absolute value is greater than the pressure deviation threshold, it indicates that the actual pressure value is lower than the target braking pressure value, and the pressure needs to be increased; the solenoid valve is then set to enter a pressurization state. When the absolute value of the pressure deviation is less than or equal to the pressure deviation threshold, it indicates that the actual pressure value is close to the target value, and the decision solenoid valve enters the pressure-holding state to maintain the current pressure. In the example of the left front wheel, the pressure deviation value is positive 0.7 MPa, and its absolute value of 0.7 MPa is greater than the threshold of 0.5 MPa. Therefore, the decision is made to close the inlet solenoid valve and open the outlet solenoid valve of the left front wheel to reduce pressure.

[0038] In some embodiments, the determined solenoid valve operating state is encoded as a specific pulse width modulation (PWM) signal, which is then sent as a brake pressure adjustment command to the solenoid valve actuator of the corresponding wheel. There is a correspondence between the solenoid valve's operating state and the duty cycle of the PWM signal. For the decompression state, a PWM signal with a specific duty cycle can be output to control the switching frequency of the discharge solenoid valve. The duty cycle of the PWM signal... Based on pressure deviation value The magnitudes of can be linearly or nonlinearly mapped, and their relationship can be expressed by the following formula: ; in: This indicates the duty cycle of the pulse width modulation signal output to the solenoid valve driver; This indicates the maximum allowed duty cycle. This represents the proportionality factor that maps the pressure deviation value to the duty cycle. This represents the calculated pressure deviation value, which is the difference between the actual pressure value and the target braking pressure value. This represents the absolute value of the pressure deviation. It can be understood that, according to the above formula, the larger the absolute value of the pressure deviation, the larger the duty cycle of the generated pulse width modulation signal, thus driving the solenoid valve to regulate the pressure with stronger action (longer opening time or higher frequency). For the aforementioned example of the left front wheel, the absolute value of the pressure deviation... The value is 0.7 MPa, assuming a proportionality coefficient. If the value is 20 MPa, then the duty cycle can be calculated. At 14%, this duty cycle signal is sent to the driver of the left front wheel outlet solenoid valve.

[0039] Optionally, the solenoid valve's opening and closing can be controlled by a solenoid valve actuator to regulate the brake fluid flow. In depressurization mode, the solenoid valve actuator periodically opens the outlet solenoid valve based on the received pulse width modulation signal, allowing some brake fluid in the brake caliper to flow back to the low-pressure reservoir. The actual pressure value of the left front wheel brake caliper begins to decrease from 9.2 MPa. The system continuously collects new actual pressure values ​​from the left front wheel brake caliper. Assuming the actual pressure value collected in the next control cycle is 8.8 MPa, the pressure deviation is recalculated as 8.8 MPa minus 8.5 MPa, equaling +0.3 MPa. Since the absolute value of the pressure deviation of 0.3 MPa is less than the threshold of 0.5 MPa, the system decides to switch the solenoid valve to pressure-holding mode, i.e., closing both the inlet and outlet solenoid valves. The pulse width modulation signal duty cycle output is zero, stabilizing the actual pressure value around 8.8 MPa. It can be understood that through a closed-loop process of real-time comparison, decision-making, and control command generation, the precise action of the solenoid valve continuously brings the actual pressure value closer to the target braking pressure value. In some embodiments, different proportional coefficients can be used for the pressurization state and the depressurization state. The mapping relationship is adapted to accommodate the different hydraulic characteristics of the braking system during pressurization and depressurization. The solenoid valve actuator can integrate current feedback to ensure that the solenoid valve coil generates accurate electromagnetic force according to the pulse width modulation signal command.

[0040] See Figure 5 The solid line in the graph represents the left front wheel slip ratio before correction. Its fluctuation range is large, with a peak slip ratio of approximately 0.23 and a trough as low as approximately 0.07, exhibiting violent periodic oscillations. This reflects the excessive response of the wheel slip ratio to changes in road surface adhesion without deceleration correction, which can easily lead to overly aggressive braking pressure adjustment, increasing the risk of wheel lock-up or vehicle instability. The dashed line represents the left front wheel slip ratio after correction. Its fluctuation range is effectively narrowed, with the peak slip ratio stabilizing at approximately 0.17 and the trough value remaining at approximately 0.13. The overall curve is smoother and maintains a similar phase relationship to the curve before correction. This dynamic correction effect stems from the dynamic adjustment of the real-time slip ratio based on the estimated longitudinal deceleration in the braking pressure decision model: when the estimated longitudinal deceleration is large (indicating good road adhesion), the slip ratio correction coefficient is small, limiting the correction range of the real-time slip ratio to fully utilize the road adhesion potential; when the estimated longitudinal deceleration is small (indicating poor road adhesion), the slip ratio correction coefficient is increased, significantly suppressing the fluctuation range of the slip ratio and ensuring that the wheel is always in a stable slip range. Simultaneously, the real-time wheel speed difference ratio between coaxial wheels serves as an auxiliary input, further balancing the braking pressure distribution between the left and right wheels and avoiding vehicle attitude deviation caused by differences in adhesion on one side.

[0041] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback, characterized in that, include: The real-time rotational speed of each wheel of the vehicle is obtained, and the instantaneous wheel speed of each wheel is calculated. The reference speed of the vehicle as a whole is calculated in real time based on the instantaneous wheel speed of each wheel. The instantaneous wheel speed of each wheel is compared with the reference vehicle speed to calculate the real-time slip ratio of the corresponding wheel. Simultaneously acquire actual longitudinal deceleration signals from the vehicle's longitudinal acceleration sensor; The actual longitudinal deceleration signal is input into the recursive observer to calculate the estimated longitudinal deceleration of the vehicle's center of gravity. Based on the real-time slip ratio of each wheel, the wheel speed difference between the left and right wheels on the same axle is calculated, and the real-time wheel speed difference ratio between the wheels on the same axle is further calculated. The estimated longitudinal deceleration, the real-time slip ratio of each wheel, and the real-time wheel speed difference ratio between wheels on the same axle are all input into the braking pressure decision model. Within the braking pressure decision model, the real-time slip ratio is dynamically corrected based on the estimated longitudinal deceleration to obtain a corrected slip ratio adapted to the current road conditions. Within the braking pressure decision model, the target braking pressure value for each wheel is jointly determined by combining the corrected slip ratio and the real-time wheel speed difference ratio. Based on the target braking pressure value of each wheel, an independent braking pressure adjustment command is generated to control the corresponding solenoid valve to adjust the actual pressure of each wheel's brake caliper. The steps for constructing a braking pressure decision model include: The actual braking process data of the vehicle under different road surface adhesion coefficients and different braking intensities are collected. The actual braking process data includes wheel speed signals, vehicle longitudinal deceleration signals and corresponding brake caliper pressure values. Feature extraction is performed on the collected actual braking process data to obtain the wheel speed change curve, slip ratio change curve, and coaxial wheel speed difference ratio curve for each working condition; The estimated longitudinal deceleration of the vehicle under each working condition is used as the input variable, and the slip ratio range corresponding to maintaining the wheel in the stable region during braking is used as the output label to form the first training dataset. The wheel speed difference ratio under each working condition is used as the input variable, and the direction and magnitude of the left and right wheel pressure difference required to maintain vehicle stability are used as the output labels to form the second training dataset. Based on the first training dataset, a slip ratio dynamic correction sub-network model is trained. Based on the second training dataset, a wheel speed difference balance sub-network model was trained. The slip ratio dynamic correction subnetwork model and the wheel speed difference balance subnetwork model are connected in parallel, and the output layers of the two subnetworks are merged into a common pressure decision fully connected layer to form a complete braking pressure decision model. The trained braking pressure decision model is stored in the storage medium of the vehicle's electronic control unit.

2. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 1, characterized in that, The process of calculating the overall reference vehicle speed in real time based on the instantaneous wheel speed of each wheel includes: The maximum value among the instantaneous wheel speeds of the vehicle's left front wheel, right front wheel, left rear wheel, and right rear wheel is selected as the initial reference wheel speed. The reference wheel acceleration is obtained by performing a time-based differential operation on the initial reference wheel speed; The acceleration of the reference wheel is compared with a preset acceleration threshold. When the acceleration of the reference wheel is greater than the acceleration threshold, it is determined that the reference wheel is in a state of violent deceleration. During the period when the reference wheel is in a state of severe deceleration, the effective reference vehicle speed calculated in the previous control cycle is used, and the longitudinal deceleration of the vehicle in the previous control cycle is combined to extrapolate and calculate the reference vehicle speed for the current control cycle. When the reference wheel is not in a state of severe deceleration, the initial reference wheel speed is directly used as the reference vehicle speed for the current control cycle.

3. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 2, characterized in that, The actual longitudinal deceleration signal is input into the recursive observer to calculate the estimated longitudinal deceleration of the vehicle's center of gravity, including: Establish a simplified longitudinal dynamics model of the vehicle that includes vehicle mass, center of gravity height, wheelbase, and wheel rolling radius; In the recursive observer, the actual longitudinal deceleration signal is used as the input excitation; Based on a simplified longitudinal dynamics model of the vehicle, state equations and observation equations are constructed in a recursive observer, where the state variables include the longitudinal velocity and longitudinal deceleration of the vehicle's center of mass. Using the algorithm of the recursive observer, the longitudinal velocity and longitudinal deceleration of the vehicle's center of mass in the current cycle are calculated by combining the state estimate of the previous cycle with the actual longitudinal deceleration signal input in the current cycle. The output longitudinal deceleration is the estimated longitudinal deceleration.

4. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 3, characterized in that, The method involves calculating the wheel speed difference between the left and right wheels on the same axle based on the real-time slip ratio of each wheel, and further calculating the real-time wheel speed difference ratio between the wheels on the same axle, including: For the same axle of the vehicle, read the instantaneous wheel speed of the left wheel and the instantaneous wheel speed of the right wheel; Calculate the absolute value of the difference between the instantaneous wheel speed of the left wheel and the instantaneous wheel speed of the right wheel to obtain the original value of the wheel speed difference of the wheels on the same axle; Read the instantaneous wheel speed of the wheel with the higher wheel speed in the coaxial section as the wheel speed reference; Divide the original value of the wheel speed difference between coaxial wheels by the wheel speed reference to obtain the normalized wheel speed difference ratio. This ratio is the real-time wheel speed difference ratio between coaxial wheels.

5. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 4, characterized in that, Within the braking pressure decision model, the real-time slip ratio is dynamically corrected based on the estimated longitudinal deceleration to obtain a corrected slip ratio adapted to the current road conditions, including: Within the braking pressure decision model, an ideal slip ratio range under standard road conditions is preset; The estimated longitudinal deceleration is matched with the pre-stored typical deceleration characteristic maps under different road surface adhesion coefficients to infer the equivalent adhesion level of the current road surface. Based on the inferred equivalent adhesion level, the upper and lower boundaries of the ideal slip ratio range are dynamically adjusted to generate the target slip ratio band adapted to the current road surface. The slip ratio of the wheel is compared with the center value of the target slip ratio band to calculate the slip ratio deviation; Based on the slip ratio deviation, a correction function related to the estimated longitudinal deceleration is applied to the real-time slip ratio to output the corrected slip ratio.

6. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 5, characterized in that, Based on the slip ratio deviation, a correction function related to the estimated longitudinal deceleration is applied to the real-time slip ratio, outputting the corrected slip ratio, including: Define a slip ratio correction factor, which is a monotonically decreasing function for estimating longitudinal deceleration; When the estimated longitudinal deceleration is large, it indicates that the road surface has good adhesion, the slip ratio correction coefficient is small, and the correction range of the real-time slip ratio is small. When the estimated longitudinal deceleration is small, it indicates poor road adhesion. A large value for the slip ratio correction coefficient results in a large correction range for the real-time slip ratio. Multiply the slip ratio deviation by the slip ratio correction factor to obtain the correction amount. Add the correction amount to the real-time slip ratio to obtain the corrected slip ratio.

7. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 6, characterized in that, Within the braking pressure decision model, by combining the corrected slip ratio and the real-time wheel speed difference ratio, an independent target braking pressure value for each wheel is jointly determined, including: A base pressure value is set for each wheel, and the base pressure value is determined by the driver's brake pedal travel. A pressure adjustment factor is preset for each wheel, with an initial value of zero. The corrected slip ratio of the wheel is compared with the preset slip ratio threshold, and the pressure adjustment factor is updated based on the comparison result. The real-time wheel speed difference ratio between coaxial wheels is compared with the preset wheel speed difference threshold, and the pressure adjustment factor is updated based on the comparison result. The target braking pressure value for the wheel is generated by weighting the pressure adjustment factor, which has been updated twice, with the base pressure value.

8. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 7, characterized in that, The step of comparing the corrected slip ratio of the wheel with a preset slip ratio threshold and updating the pressure adjustment factor based on the comparison result includes: Preset high and low slip ratio thresholds for each wheel; When the corrected slip ratio of the wheel is greater than the slip ratio high threshold, the wheel is judged to be slipping excessively, and a negative adjustment amount is added to the pressure adjustment factor. When the corrected slip ratio of the wheel is less than the low slip ratio threshold, the wheel is judged to be underslip, and a positive adjustment amount is added to the pressure adjustment factor. When the wheel's corrected slip ratio is between the low slip ratio threshold and the high slip ratio threshold, the pressure adjustment factor is not updated based on the slip ratio.

9. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 8, characterized in that, The step of comparing the real-time wheel speed difference ratio between coaxial wheels with a preset wheel speed difference threshold and updating the pressure adjustment factor based on the comparison result includes: A threshold value for the wheel speed difference ratio is preset for the left and right wheels on the same axle; Compare the real-time wheel speed difference ratio of the left and right wheels on the same axle with the wheel speed difference ratio threshold; When the real-time wheel speed difference ratio is greater than the wheel speed difference ratio threshold, it is determined that the difference in wheel adhesion on both sides of the axle is significant. A positive adjustment is added to the pressure adjustment factor of the wheel on the side with lower wheel speed, and a negative adjustment is added to the pressure adjustment factor of the wheel on the side with higher wheel speed, or its pressure adjustment factor is kept unchanged. When the real-time wheel speed difference ratio is less than or equal to the wheel speed difference ratio threshold, the pressure adjustment factor of any wheel on the axle is not updated based on the wheel speed difference ratio.

10. The vehicle anti-lock braking control method based on wheel speed difference and deceleration feedback according to claim 9, characterized in that, The process of generating independent brake pressure adjustment commands based on the target brake pressure value for each wheel, and controlling the corresponding solenoid valves to adjust the actual pressure of each wheel's brake caliper, includes: Real-time acquisition of the actual pressure value of each wheel brake caliper; Compare the actual pressure value of each wheel with the target braking pressure value of the corresponding wheel, and calculate the pressure deviation value; Based on the sign and magnitude of the pressure deviation, the working state of the corresponding solenoid valve is determined. The working states include pressure boosting, pressure holding, and pressure reducing. The determined solenoid valve operating state is encoded into a specific pulse width modulation signal, which is then sent as a brake pressure adjustment command to the solenoid valve driver of the corresponding wheel. The solenoid valve is controlled by a solenoid valve actuator to open and close the solenoid valve, thereby adjusting the brake fluid flow rate and bringing the actual pressure value closer to the target brake pressure value.

Images