Vehicle slipping out-of-trouble control method, device and equipment and storage medium

Abstract

The invention relates to the technical field of automatic driving, and discloses a vehicle slipping escape control method, device and equipment and a storage medium. According to the method, driving data are collected in real time through a multi-mode sensor, a slipping judgment index is calculated, and a self-adaptive escape strategy is dynamically generated in combination with road surface information; the problems that a traditional method is inaccurate in slip judgment and lags behind a control strategy are solved, and the method has the advantage that the escape efficiency and safety are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a vehicle skid escape control method, device, equipment and storage medium. BACKGROUND

[0002] In the process of driving on complex terrain or low adhesion road surface, the vehicle is prone to skid phenomenon, which leads to the vehicle being trapped in trouble, affecting the traffic efficiency and driving safety.

[0003] Especially for the automatic driving vehicle, in the vehicle only existing traction control system and vehicle stability control system, the control developer mainly improves the success rate of escape by the comprehensive operation of the throttle and gear shift according to the driving experience and test scene, but no matter how to test, the scene is limited, and these systems still have problems such as response lag and untimely adaptive adjustment when facing serious skid or trapped vehicle situation. Therefore, there is an urgent need for a control method which can automatically distinguish the skid situation based on the real-time state of the vehicle and actively take effective escape strategy. SUMMARY

[0004] The present application provides a vehicle skid escape control method, device, equipment and storage medium, which is used to solve the problem of untimely control and low success rate of existing escape solution.

[0005] The first aspect of the present application provides a vehicle skid escape control method, comprising: Collecting the current driving data of the vehicle through the multi-modal sensor arranged on the vehicle, wherein the driving data includes the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body posture parameter and the GPS displacement change value; Calculating the skid judgment index of each tire based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body posture parameter and the GPS displacement change value; After judging the tire skid based on the skid judgment index, controlling the vehicle to enter the skid escape mode, and collecting the road surface information in front of the vehicle and the tire area; Generating the corresponding escape control strategy based on the road surface information and the skid judgment index, and controlling the vehicle to drive for escape.

[0006] Further, the calculation of the skid judgment index of each tire based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body posture parameter and the GPS displacement change value comprises: Preliminary judging whether the vehicle is skidding based on the vehicle body posture parameter and the GPS displacement change value; If yes, a slip determination index of each wheel is calculated based on the rotational speed of each wheel, the acceleration of the vehicle, and the steering wheel angle.

[0007] Further, after the slip determination index of each wheel is calculated based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body posture parameter, and the GPS displacement change value, the method further comprises: determining whether the slip determination index reaches a preset index value; If yes, the wheel speed difference, the vehicle longitudinal acceleration change rate, and the vehicle body pitch angle are calculated, and a slip judgment function is constructed based on the wheel speed difference, the vehicle longitudinal acceleration change rate, and the vehicle body pitch angle; solving the slip judgment function to determine whether the vehicle is in a slip state; If yes, it is determined that the tire is slipping.

[0008] Further, the road surface information in the front of the vehicle and the tire area is collected, comprising: acquiring a three-dimensional terrain image of the front of the vehicle and the tire contact area through a laser radar, a camera, or an ultrasonic sensor; performing grid processing on the three-dimensional terrain image to extract road surface information of the area corresponding to each wheel, wherein the road surface information includes the current depth of the wheel position, the slope, and the road surface adhesion coefficient.

[0009] Further, the corresponding escape control strategy is generated based on the road surface information and the slip determination index, and the vehicle is controlled to perform escape driving, comprising: calculating the driving force distribution ratio of each wheel according to the slip determination index of each wheel and the corresponding slope and road surface adhesion coefficient; comparing the road surface adhesion coefficient of each wheel with a pre-set driving force threshold value; based on the comparison result, optimizing and adjusting the driving force distribution ratio of each wheel, wherein the optimization and adjustment includes reducing the driving force ratio of the wheel with a road surface adhesion coefficient lower than the driving force threshold value and increasing the driving force ratio of the wheel with a road surface adhesion coefficient higher than the driving force threshold value; dynamically adjusting the throttle opening and the steering wheel angle of the vehicle according to the optimized driving force ratio, and outputting the corresponding escape control strategy to control the vehicle to perform escape driving.

[0010] Further, the dynamically adjusting the throttle opening and the steering wheel angle of the vehicle according to the optimized driving force ratio, and outputting the corresponding escape control strategy to control the vehicle to perform escape driving, comprising: input the optimized driving force ratio, the current throttle opening of the vehicle, and the current depth of the wheel position, the road adhesion coefficient, and the vehicle attitude parameter into a pre-trained combination model of a multilayer perception and a recurrent neural network to output a signal control sequence for throttle opening and steering wheel angle control; The signal control sequence is executed to control the vehicle to perform the escape driving.

[0011] Further, after the signal control sequence is executed to control the vehicle to perform the escape driving, the method further comprises: real-time monitoring of the vertical load of each wheel and the real-time depth of the wheel position during the escape process; If the vertical load of a certain wheel is not greater than a set value and the real-time depth is greater than the current depth, the suspension height of the corresponding wheel is lowered to increase the ground contact area, and the escape driving is continued; If the vertical load of a certain wheel is greater than a set value, the suspension height of the corresponding wheel is raised until the vehicle cabin is in a horizontal position balance state, and the escape driving is continued.

[0012] The second aspect of the present application provides a vehicle slip escape control device, comprising: a collection module configured to collect current driving data of the vehicle through a plurality of modal sensors arranged on the vehicle, wherein the driving data includes the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameter, and the GPS displacement change value; a calculation module configured to calculate a slip determination index of each tire based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameter, and the GPS displacement change value; a control module configured to control the vehicle to enter a slip escape mode after determining tire slip based on the slip determination index, collect road surface information in front of the vehicle and in the tire area, and generate a corresponding escape control strategy based on the road surface information and the slip determination index, and control the vehicle to perform escape driving.

[0013] The third aspect of the present application provides a vehicle slip escape control device, comprising a memory and at least one processor, wherein the memory stores instructions; and the at least one processor invokes the instructions in the memory to enable the vehicle slip escape control device to perform the vehicle slip escape control method described above.

[0014] The fourth aspect of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores instructions, and when the instructions are executed on a computer, the computer executes the vehicle slip escape control method described above.

[0015] In the technical scheme provided in the application, the current driving data of the vehicle is collected by the multi-modal sensor arranged on the vehicle, the driving data including the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the body posture parameter and the GPS displacement change value; the slip determination index of each tire is calculated based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the body posture parameter and the GPS displacement change value; after the slip determination index is used to determine that the tire slips, the vehicle is controlled to enter the slip escape mode, and the road surface information in front of the vehicle and the tire area is collected; the corresponding escape control strategy is generated based on the road surface information and the slip determination index, and the vehicle is controlled to drive for escape. The application realizes real-time detection of the slip state by collecting the driving state of the vehicle and the road surface information in real time through the multi-modal sensor, combining the slip determination index calculation formula and the artificial intelligence dynamic control strategy, so as to ensure the real-time performance of the escape control and the success rate of the vehicle escape. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The first flowchart of the vehicle slip escape control method provided in the application is shown in the figure. Figure 2 The second flowchart of the vehicle slip escape control method provided in the application is shown in the figure. Figure 3 The structure diagram of the vehicle slip escape control device provided in the application is shown in the figure. Figure 4 The structure diagram of the vehicle slip escape control device provided in the application is shown in the figure. DETAILED DESCRIPTION

[0017] The terms "first", "second", "third", "fourth" and the like in the specification and claims of the application and in the above description of the drawings (if any) are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein.

[0018] In addition, the term "comprising" or "having" and any variation thereof is intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0019] In the related art, the automatic driving vehicle transportation operation in a complex working condition environment faces many challenges. Especially when driving in the rain or a humid environment, the vehicle often slips and cannot move forward. The existing slip control method mainly relies on a single sensor signal (such as a wheel speed sensor) to determine the slip state, and realizes partial slip suppression through simple braking or engine torque control. However, the current control method cannot accurately identify the slip reason in a complex terrain, cannot accurately determine the reason, and cannot give the corresponding control scheme. Therefore, the present application proposes a fine slip determination combined with real-time road information by fusing multi-modal sensors, and outputs the corresponding escape control and strategy to realize higher precision and timely slip detection and higher success rate of escape control.

[0020] To solve the above problems, the specific process of the present application is described below, referring to Figure 1 For the first embodiment of the vehicle slip escape control method of the present application, the vehicle slip escape control method comprises the following steps: 101. Collect the current driving data of the vehicle through the multi-modal sensors arranged on the vehicle, which includes the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the body posture parameter and the GPS displacement change value; Specifically, the multi-modal sensors can include a wheel speed sensor, an acceleration sensor, a steering wheel angle sensor, a body posture sensor and a GPS module. All sensors transmit data to the central control unit (VCU) through the CAN bus, and the sampling frequency is preferably above 50Hz to ensure the real-time response.

[0021] Among them, the wheel speed sensor is installed at the hub position of the four wheels for real-time detection of the rotation speed ωi (unit: rad / s) of each wheel; The acceleration sensor is arranged at the position close to the chassis in the driver's cabin of the vehicle for detecting the longitudinal, lateral and vertical accelerations a x , a y , a z (unit: m / s²) of the vehicle; The steering wheel angle sensor is installed on the steering shaft of the vehicle steering wheel for detecting the current steering wheel angle δ (unit: °) to reflect the steering intention of each steering wheel adjustment; The body posture sensor (IMU) is installed on the bottom of the vehicle compartment or the suspension of the wheels for detecting and outputting the pitch angle θp, roll angle θr and yaw angle θy of the vehicle body; The GPS module is installed in the driver's cabin of the vehicle for collecting the position coordinates (X, Y) of the vehicle and the unit time displacement change ΔS (unit: m / s).

[0022] 102、calculate a slip determination index of each tire based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the body posture parameter and the GPS displacement change value; In this step, the slip determination index can be calculated by fusing the wheel speed difference, the acceleration change rate and the body posture parameter. Further, the ratio of the wheel speed difference to the theoretical slip rate, and the correlation function of the longitudinal acceleration in the acceleration and the steering wheel angle can be calculated, and the ratio, the correlation function and the body posture parameter are fused based on the preset weight to obtain the slip determination index.

[0023] 103、after determining the tire slip based on the slip determination index, control the vehicle to enter the slip escape mode, and collect the road surface information in front of the vehicle and the tire area; Specifically, whether the slip determination index reaches a preset index value is judged. If yes, the wheel speed difference, the vehicle longitudinal acceleration change rate and the body pitch angle are calculated, and a slip judgment function is constructed based on the wheel speed difference, the vehicle longitudinal acceleration change rate and the body pitch angle. The slip judgment function is solved to determine whether the vehicle is in a slip state. If yes, it is determined that the tire is slipping.

[0024] It should be noted that the preset index value herein refers to a critical parameter preset for triggering the slip state depth analysis, which can be realized by experimental calibration or dynamic threshold algorithm. Its function is to filter incidental signal fluctuations.

[0025] Before determining whether it is slipping, the wheel speed difference, the vehicle longitudinal acceleration change rate and the body pitch angle are calculated.

[0026] The calculation of the wheel difference includes: dividing the wheel speed of each wheel collected by the wheel speed sensor into driving wheels and driven wheels, calculating the standard deviation of the wheel speed based on the wheel speed of the driving wheels and the driven wheels, and taking the standard deviation as the wheel speed difference to reflect the motion state difference between the driving wheels and the driven wheels.

[0027] The calculation of the vehicle longitudinal acceleration change rate is obtained by calculating the first order derivative of the vehicle acceleration collected by the acceleration sensor, which is used to represent the dynamic characteristics of the vehicle power output.

[0028] The calculation of the body pitch angle is specifically to solve the body posture parameter collected by the inertial measurement unit to obtain the attitude angle, and further obtain the body pitch angle based on the attitude angle.

[0029] When constructing the slip judgment function, a weighted linear combination or a nonlinear regression model is used to construct the slip judgment function. Specifically, the wheel speed difference, the vehicle longitudinal acceleration change rate and the body pitch angle are weighted and calculated according to the set weighting coefficient, and the linear regression function is used for processing to obtain the slip judgment function.

[0030] In practical applications, when the slip determination index of a certain wheel is detected to exceed the preset threshold, the system synchronously collects the wheel speed difference, longitudinal acceleration rate of change, and vehicle body pitch angle data. The wheel speed difference is calculated as the absolute value of the speed difference between the two wheels on the same shaft, the longitudinal acceleration rate of change is obtained by time series difference operation of the acceleration sensor data, and the vehicle body pitch angle is output in real time by the inertial measurement unit. The three parameters are input into the slip judgment function, which can be expressed as f(Δv, a', θ) = k1·Δv + k2·a' + k3·θ, where k1, k2, and k3 are weight coefficients calibrated according to the vehicle dynamics characteristics, Δv is the wheel speed difference, a' is the vehicle longitudinal acceleration rate of change, and θ is the vehicle body pitch angle. When the function output value exceeds the critical threshold, it is determined that the vehicle enters a substantial slip state, and the escape control module is triggered at this time.

[0031] Further, the road surface information can be understood as the three-dimensional features of the terrain obtained by a laser radar or a camera, and specifically, the recess depth, slope, and adhesion coefficient can be extracted by grid processing.

[0032] 104. Based on the road surface information and the slip determination index, a corresponding escape control strategy is generated, and the vehicle is controlled to drive out of the trouble.

[0033] In this step, the escape control strategy can include specific drive torque distribution, steering wheel angle correction, and throttle control curve, so as to output optimal vehicle control instructions when a low adhesion area or tire sinking situation is detected, and realize stable escape.

[0034] Specifically, it can be obtained based on the control instructions for dynamically adjusting the drive force distribution ratio according to the road surface adhesion coefficient in the road surface information and the slip index, for example, by adjusting the torque output of each wheel through an optimization algorithm, and combining the throttle opening and steering angle control to realize escape.

[0035] In this embodiment, when the vehicle is driving, the multi-modal sensor collects wheel speed, acceleration, steering wheel angle, and other data in real time, and judges the overall motion trend of the vehicle in combination with the GPS displacement change. The slip determination index is generated by calculating the slip rate difference of each wheel and the change of the vehicle body pitch angle. If the index exceeds the threshold, the system switches to the escape mode, starts the laser radar to scan the front road surface, and extracts the recess depth and adhesion coefficient of the area corresponding to each wheel. Based on the difference in road surface adhesion coefficient, the system redistributes the drive force of each wheel, reduces the torque output of the wheel in the low adhesion area, adjusts the steering wheel angle to optimize the driving trajectory, and finally realizes escape through differential control and power adjustment.

[0036] By fusing multi-dimensional sensor data and real-time road scanning, for example, in a sand pit vehicle scene, the prior art can continuously increase the throttle to cause the wheels to spin, while the present solution can actively reduce the driving force of the spinning wheels according to the real-time depth of the wheel position and the adhesion coefficient, and increase the output of the high adhesion wheels, thereby quickly escaping from the pit.

[0037] Through the above technical solution, the vehicle state and road surface characteristics can be perceived in real time, and an escape strategy that adapts to the current scene can be dynamically generated, thereby realizing accurate judgment of the slipping state and dynamic strategy generation, solving the problems of response lag and single adjustment, and significantly improving the escape efficiency and success rate in complex terrain.

[0038] Referring to Figure 2 A second vehicle slipping escape control method is provided for the embodiments of the present application. By introducing multi-modal sensor fusion technology, multi-dimensional data such as wheel speed, vehicle attitude, acceleration, etc. can be synchronously acquired, and a comprehensive judgment model can be constructed in combination with road three-dimensional information, so as to optimize the driving force distribution and vehicle control strategy. Specifically, the following steps are included: 201. Collecting the current driving data of the vehicle through the multi-modal sensors arranged on the vehicle, the driving data including the speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle attitude parameters, and the GPS displacement change value.

[0039] The multi-modal sensor should be understood as a combination device integrating wheel speed sensors, accelerometers, gyroscopes, GPS modules, etc. Specifically, wheel speed sensors can be used to measure wheel speed, inertial measurement units can be used to collect vehicle attitude parameters, and GPS modules can be used to obtain displacement change values, which are used to comprehensively reflect the vehicle motion state.

[0040] 202. Preliminarily judging whether the vehicle is slipping based on the vehicle attitude parameters and the GPS displacement change value. Specifically, after the GPS displacement change values at consecutive multiple time points are the same, it is determined whether the change rule of the vehicle attitude parameters at the multiple time points is the same as or similar to a preset slipping change rule. If so, it is determined that the vehicle is suspected to be slipping.

[0041] 203. If so, the slipping judgment index is calculated based on the speed of each wheel, the acceleration of the vehicle, and the steering wheel angle using the calculation formula of the slipping judgment index. It should be noted that the calculation formula of the slipping judgment index can fuse the wheel speed, acceleration, and steering parameters into a quantitative index through a mathematical relationship, such as weighted summation or a nonlinear function model. Specifically, it can be Sk= (ω*r-v) / v, where ω is the wheel angular velocity determined based on the wheel speed and the steering wheel angle, r is the radius of the wheel, and v is the vehicle speed calculated based on the acceleration of the vehicle.

[0042] The vehicle body posture parameters are obtained by measuring the roll angle, pitch angle and yaw angle of the vehicle using a gyroscope and an acceleration sensor. The GPS displacement change value can be understood as the trajectory data of the change of the vehicle position with time, which can be obtained in real time through differential GPS technology.

[0043] In actual application, when the vehicle is driving, the gyroscope and the acceleration sensor continuously collect the vehicle body posture parameters, and the differential GPS module updates the displacement change value at a fixed frequency. By comparing the corresponding relationship between the change trend of the vehicle body posture angle and the GPS displacement, if the posture fluctuates dramatically but the displacement increment is significantly lower than expected, the preliminary slip judgment is triggered. After confirming the preliminary slip, the system calls the preset calculation formula, such as fusing the deviation value of the wheel speed and the theoretical speed, the attenuation coefficient of the longitudinal acceleration and the correction factor of the steering wheel angle to generate an independent slip determination index for each wheel. The index can be quantified as a value in the range of 0-1, and the higher the value, the greater the risk of slip.

[0044] Here, through the dual verification mechanism of fusing the vehicle body posture and the GPS displacement, the real slip is effectively distinguished from the normal driving disturbance, and the calculation formula of the multi-dimensional dynamic parameter is combined to make the slip determination index have higher environmental adaptability and accuracy.

[0045] 204, judging whether the wheel is a real slip according to the slip determination index; Specifically, whether the slip determination index reaches a preset index value is judged; if it does, the wheel speed difference, the vehicle longitudinal acceleration change rate and the vehicle body pitch angle are calculated, and a slip judgment function is constructed based on the wheel speed difference, the vehicle longitudinal acceleration change rate and the vehicle body pitch angle; the slip judgment function is solved to judge whether the vehicle has a slip state; if it does, the tire slip is determined.

[0046] 205, controlling the vehicle to enter a slip escape mode, and collecting road surface information in front of the vehicle and the tire area; In this embodiment, a laser radar, a camera or an ultrasonic sensor is used to obtain a three-dimensional terrain image in front of the vehicle and a tire contact area; the three-dimensional terrain image is subjected to grid processing to extract road surface information of the area corresponding to each wheel, wherein the road surface information includes the current depth of the concave, the slope and the road adhesion coefficient of the wheel position.

[0047] It should be noted that the laser radar, the camera or the ultrasonic sensor refers to a sensing device for obtaining three-dimensional space information, which can be realized in a multi-modal fusion manner to improve the robustness of data acquisition through the complementary characteristics of different sensors.

[0048] Grid processing refers to dividing the three-dimensional terrain image into uniformly distributed unit grids, which can be realized by image segmentation algorithm, and the micro-terrain features of the wheel contact surface are extracted through local area analysis.

[0049] The current depression depth refers to the amount of subsidence of the ground under the wheel, which can be calculated by three-dimensional point cloud data, and is used to evaluate the degree of wheel sinking.

[0050] The slope refers to the inclination angle of the wheel contact surface with the horizontal plane, which can be calculated by normal vector analysis algorithm, and is used to judge the lateral force of the terrain on the wheel.

[0051] The road adhesion coefficient refers to the friction characteristic parameter between the tire and the ground, which can be estimated by texture recognition and historical data matching model, and is used to predict the slip trend of the wheel.

[0052] In practical application, after detecting vehicle slip, three-dimensional terrain data of the front of the vehicle and the tire contact area are synchronously collected by multiple types of sensors. For example, a laser radar can generate high-precision point cloud data, a camera can capture texture information, and an ultrasonic sensor can supplement near-distance obstacle information. Then, the fused three-dimensional image is segmented into several grid units, each unit corresponding to the local area of the wheel ground. For the contact area of each wheel, the depression depth, slope and adhesion coefficient parameters of the wheel position are extracted respectively. The depression depth is obtained by calculating the vertical distance of the ground points in the grid unit from the reference plane, the slope is determined by fitting the normal vector direction of the ground points in the grid unit, and the adhesion coefficient is estimated by combining the texture features of the grid unit and the pre-trained road friction model.

[0053] 206、Based on the road surface information and the slip determination index, a corresponding escape control strategy is generated, and the vehicle is controlled to drive out of the trouble.

[0054] Specifically, according to the slip determination index and the corresponding slope and road adhesion coefficient of each wheel, the driving force distribution ratio of each wheel is calculated; the road adhesion coefficient of each wheel is compared with the pre-set driving force threshold; based on the comparison result, the driving force distribution ratio of each wheel is optimized and adjusted, wherein the optimization and adjustment includes reducing the driving force ratio of the wheel with road adhesion coefficient lower than the driving force threshold and increasing the driving force ratio of the wheel with road adhesion coefficient higher than the driving force threshold; the throttle opening and steering wheel angle of the vehicle are dynamically adjusted according to the optimized driving force ratio, and the corresponding escape control strategy is output to control the vehicle to drive out of the trouble. Thus, the power output of each wheel can be accurately adjusted according to the real-time road surface state, avoiding energy waste and tire wear caused by continuous application of driving force in low adhesion area, while the traction efficiency of the wheel in high adhesion area is enhanced through differential control, significantly improving the escape success rate of the vehicle in complex terrain.

[0055] It should be noted that the driving force distribution ratio refers to the ratio of the required output power of each wheel to the total driving force, which can specifically use the product of the wheel speed difference and the road adhesion coefficient as a calculation parameter, and obtain the ratio value through normalization processing.

[0056] The driving force threshold refers to a pre-set road adhesion coefficient critical value, which can be an empirical value between 0.3 and 0.5, for example, for determining whether the wheel is in a state of easy slipping.

[0057] The optimization adjustment refers to dynamically adjusting the power output weight according to the comparison result of the adhesion coefficient and the threshold value, for example, when the adhesion coefficient of a certain wheel is lower than the threshold value, the driving force ratio is reduced to 60%-80% of the original value, and the adhesion coefficient of the wheel is higher, the ratio is increased to 120%-150%.

[0058] In this embodiment, after the vehicle enters the slipping escape mode, the laser radar obtains the depth and slope data of the wheel position, and the road adhesion coefficient extracted by the camera is fused, combined with the real-time calculation of the slipping determination index of each wheel, and a power distribution model is established. For example, when the adhesion coefficient of the left front wheel is 0.4 and higher than the threshold value 0.35, the driving force ratio is increased to 1.3 times, and the adhesion coefficient of the right rear wheel is 0.28, which is lower than the threshold value, and the ratio is adjusted to 0.7 times. The adjusted ratio is converted into throttle opening command by the vehicle control unit, for example, the total driving force is distributed to each driving motor in proportion, and the steering angle is corrected combined with the steering wheel angle data, so that the vehicle reduces the power output in the low adhesion area and enhances the traction in the high adhesion area.

[0059] In another embodiment, the driving force ratio is dynamically adjusted according to the optimized driving force ratio, the throttle opening and the steering wheel angle of the vehicle, and the corresponding escape control strategy is output to control the vehicle to drive in the escape mode, comprising: The optimized driving force ratio, the current throttle opening of the vehicle, and the current depth of the wheel position, the road adhesion coefficient, and the vehicle attitude parameter are input into the combined model of the pre-trained multi-layer perception and recurrent neural network to output the signal control sequence of the throttle opening and the steering wheel angle control; executing the signal control sequence controls the vehicle to drive in the escape mode.

[0060] It should be noted that the combined model of the multi-layer perception and the recurrent neural network refers to a machine learning model in which the multi-layer perception module processes static feature data and the recurrent neural network module processes time series dynamic data. Specifically, it can be realized by combining offline training and online fine-tuning, for example, through supervised learning of sensor data and successful escape control signal sequences in historical escape scenarios, so that the model can capture the dynamic response law under different road conditions.

[0061] The signal control sequence refers to a set of accelerator opening degree and steering wheel corner adjustment instructions arranged in chronological order, which can be implemented by control amount sequence in discrete time steps, for example, generating a control instruction sequence for the next 2 seconds at intervals of 0.1 seconds to ensure the continuity of vehicle dynamic adjustment.

[0062] Further, the accelerator control signal sequence is converted into a control strategy, specifically: Initialize the accelerator opening degree parameter, and estimate the initial value according to Sk and the initial value of acceleration; Construct an AI accelerator control model F(x) trained based on historical driving trajectory and wheel state data, x is an input feature vector, and the initial value * is the optimal accelerator opening degree at the current time; The initial value * is transmitted to the electronic accelerator module as a control command to achieve dynamic adjustment of the accelerator and optimize the escape.

[0063] The input feature vector is calculated by inputting the current tire slip determination index Sk, the depth of the wheel position, the vehicle longitudinal acceleration, the vehicle inclination angle, the steering wheel corner, the vehicle speed, and the historical escape times into the multi-layer perception neural network.

[0064] After the vehicle enters the escape mode, the optimized driving force ratio and the real-time collected road surface parameters and vehicle state parameters are synchronously input into the combined model. The multi-layer perception performs nonlinear mapping on the current depth of the wheel position and the road adhesion coefficient, and the recurrent neural network predicts the future state change trend based on the time series data of the vehicle attitude parameters. The outputs of the two modules are combined to generate a control signal sequence of the accelerator opening degree and the steering wheel corner, such as the superposition result of the feedforward control amount and the feedback correction amount. During execution, the vehicle controller executes each control instruction in time steps, and simultaneously optimizes the subsequent instructions based on real-time sensor data, for example, when the laser radar detects a new road depression, the model is triggered to recalculate the remaining control sequence.

[0065] 207、Real-time monitoring of the parameters of each wheel during the escape process, and real-time adjustment of the escape control strategy based on the monitored parameters until the escape is completed.

[0066] In this step, the vertical load of each wheel and the real-time depth of the wheel position during the escape process are monitored in real time; if the vertical load of a wheel is not greater than a set value and the real-time depth of the wheel position is greater than the current depth, the suspension height of the corresponding wheel is reduced to increase the ground contact area, and the escape driving continues; if the vertical load of a wheel is greater than a set value, the suspension height of the corresponding wheel is increased until the vehicle cabin is in a horizontal position and the balance state is achieved, and the escape driving continues.

[0067] It should be noted that the vertical load refers to the vertical pressure borne by the tire-ground contact surface during vehicle driving, which can be measured by a pressure sensor at the hub bearing, and is used to determine whether the tire is abnormally pressed due to suspension deformation.

[0068] The real-time depression depth refers to the instantaneous sinking amount of the position of the tire on the ground, which can be scanned by a laser radar or an ultrasonic sensor, and is used to evaluate the degree of sinking of the tire into soft road surface.

[0069] The suspension height adjustment refers to changing the stroke of the suspension system by an electrically controlled hydraulic or pneumatic device, which can be realized by an actuator in an active suspension system, and is used to dynamically adjust the tire grounding state to enhance the grip.

[0070] During the execution of the vehicle escape driving instruction, the vertical load data of each tire is continuously collected by the hub pressure sensor, and the depression depth of the tire contact area wheel position is obtained by using the terrain scanning device. When it is detected that the vertical load of a certain tire is lower than the set threshold and the ground depression at the position is continuously deepening, it indicates that the tire may be insufficient in ground contact area due to excessive compression of the suspension, at which time the suspension height is lowered to increase the contact area of the tire with the ground. On the contrary, when the vertical load of a certain tire exceeds the threshold, it indicates that the suspension system may cause excessive pressure on one side due to the tilting of the vehicle body, at which time the suspension height is raised to restore the balance of the vehicle body. The suspension adjustment process is synchronized with the escape driving to form a closed loop control. Here, through the joint judgment of the vertical load and the depression depth, the suspension adjustment can be triggered immediately when the tire starts to sink, preventing the vehicle from further sinking into the soft road surface. At the same time, through the coordinated control of the suspension height and the balance state of the vehicle body, the power distribution strategy can be optimized under the condition of the optimized tire grounding, so as to significantly improve the success rate and safety of the escape operation.

[0071] In another possible implementation, the escape control strategy includes the number of forward and reverse driving of the vehicle; and the control of the vehicle to perform escape driving includes: initializing an escape cycle counter to record the current round of forward and reverse driving of the vehicle; in each round of reverse and forward driving, recording the distance Δd1 of forward driving and the distance Δd2 of backward driving, and the change rate ΔSk of the slip determination index of each wheel before and after escape; dynamically adjusting the duration of the next round of reverse or forward driving according to the change trend of the change rate ΔSk and the distance Δd1 and the distance Δd2, and updating the cycle counter until the escape is successful or the maximum cycle number is exceeded.

[0072] Specifically, the escape cycle counter can be realized by using a register of an embedded system or an independent storage unit, which functions to quantify the number of escape operations to avoid infinite loop.

[0073] The forward and reverse wheel times are realized by counting the driving direction signal by the vehicle control unit, which adjusts the vehicle position by periodic action to escape the slipping area.

[0074] The change rate of the slipping determination index ΔSk is the numerical fluctuation amplitude of the slipping determination index of each wheel in unit time, which can be realized by differential calculation or sliding window statistical method on sensor data, and reflects the improvement degree of the wheel slip state by the escape operation.

[0075] The dynamic adjustment duration is realized by using a fuzzy logic control algorithm or a proportional integral derivative controller, which avoids invalid escape or energy waste caused by fixed time operation.

[0076] Specifically, after the vehicle enters the escape mode, the escape cycle counter is first initialized and the initial wheel time is recorded. After each forward and reverse combination operation is completed, the vehicle displacement data is obtained by the vehicle sensor and Δd1 and Δd2 are calculated, and ΔSk of each wheel is calculated based on wheel speed and other parameters. When it is detected that ΔSk presents a downward trend and the cumulative displacement of Δd1 and Δd2 does not reach the expected threshold, the control unit will shorten the duration of the next operation; if ΔSk continues to rise and the displacement change meets the expectation, the operation time is extended. The cycle counter is automatically updated after each operation, and the escape process is terminated when the cumulative cycle number reaches the preset maximum value or the slipping determination index falls below the safety threshold.

[0077] In another embodiment, the control vehicle to escape driving also includes: The terrain depression of each wheel is scanned in real time by a depth camera and a laser radar; A terrain slipping weight factor is constructed based on the depression and the slipping determination index, wherein the calculation formula of the terrain slipping weight factor is Wi=α·Sk+β·di, where α and β are empirical weight factors, and di is the depression; Based on the size of the terrain slipping weight factor, the wheels of the vehicle are respectively prioritized for escape, wherein the wheel with a larger terrain slipping weight factor has a higher priority, and the priority of the wheel is improved by differential control; The escape control strategy is regenerated based on the wheels with adjusted priority.

[0078] Specifically, the terrain depression is extracted by three-dimensional point cloud data or grid image processing technology to obtain the depression depth and edge contour information of the wheel position, for example, three-dimensional reconstruction is performed by using the ranging data of the laser radar combined with the image of the depth camera. This feature is used to quantify the severity of the wheel being trapped in the obstacle, and provides terrain parameters for weight calculation.

[0079] The terrain slip weight factor refers to the evaluation parameter of combining the slip determination index and the terrain depression degree. Specifically, the two can be combined by linear weighting, for example, a can take a value range of 0.6-0.8, and β takes a value range of 0.2-0.4, and the influence of different factors on the escape priority is reflected by adjusting the weight ratio. The factor is used to dynamically evaluate the escape emergency degree of each wheel, and solves the misjudgment problem caused by relying on a single slip index in the traditional method.

[0080] The differential control refers to changing the speed difference between the wheels by adjusting the driving torque distribution ratio of each wheel. Specifically, it can be realized by using electronic differential lock or torque vectoring system. This technical means actively adjusts the driving force of the high-priority wheel to enhance its traction efficiency in complex terrain. Specifically, different driving torques T_i are applied to the high-priority wheel, so that the slipping wheel reduces the driving force, and the non-slip wheel increases the driving force; Ti is adjusted in real time according to the wheel speed difference Δω=ωi-ωavg to ensure that the vehicle maintains the forward direction without deflection.

[0081] After detecting that the vehicle enters the slip state, three-dimensional terrain data of each wheel position is obtained by multi-sensor fusion. For example, the laser radar obtains point cloud data at a scanning frequency of 10-20 Hz, and the depth camera synchronously collects image information, and the data of the two are fused to generate the terrain depression parameter di through coordinate transformation. Then di and the real-time calculated slip determination index Sk are substituted into the weight formula, for example, when the Sk value of a certain wheel is 0.85 and di is 15 cm, under the weight of α=0.7, β=0.3, the Wi value can reach 0.7*0.85+0.3*15=0.595+4.5=5.095. According to the calculation result, the wheels are sorted, and the differential control is preferentially implemented on the wheel with the highest Wi value, for example, by the electronic control unit to allocate a higher proportion of torque output to the wheel, while reducing the driving force of the low-priority wheel.

[0082] In some specific embodiments, the calculation of the terrain depression degree can use a dynamic threshold adjustment mechanism. For example, when the vehicle is in a sandy environment, the reference value of di is set to 10 cm, and in a snowy environment, it is adjusted to 5 cm to adapt to the characteristics of different road conditions. In addition, the execution of differential control can be combined with vehicle yaw rate sensor data to adjust the steering angle while increasing the driving force of a specific wheel, preventing the vehicle from losing control due to unilateral driving force mutation.

[0083] In another embodiment, whether the vehicle is slipping can be realized by a multi-modal confirmation method, which specifically includes the following steps: Determine whether there is idling by comparing the wheel speed with the GPS speed; The body posture change measured by the gyroscope is used to determine whether there is a wheel imbalance and skid tendency; The temperature sensor, current load sensor and other information are used to determine whether the wheel is in a special scene such as mud or snow, thereby improving the accuracy of skid detection.

[0084] Further, in order to avoid excessive consumption of the vehicle, an energy saving constraint strategy is also introduced in the skid escape mode, which is used to optimize the energy consumption control of the vehicle in the escape process. Specifically: During the escape control process, the motor power consumption P_i and the change rate of ΔSk of each reverse / forward process are recorded; the unit power consumption escape efficiency index E_i = ΔSk / P_i is constructed; the energy consumption strategy of the subsequent action is dynamically optimized according to E_i to improve the overall escape energy efficiency.

[0085] Further, in order to optimize the escape control, the combination model of multi-layer perception and recurrent neural network can also refer to the online incremental learning mechanism to adjust the escape control strategy and improve the success rate of escape. Specifically, it includes: The relationship between the input x, the control result θ* and the actual feedback Sk is continuously recorded during the escape control process; New samples are cached in a fixed window manner, and the original training set is periodically retrained to update the model parameters; The ADAM optimizer is used for fine-tuning update, so that the model F(x) gradually adapts to the environment changes.

[0086] In another embodiment, a driving record data saving mechanism is triggered when the vehicle enters the escape control mode, which includes: recording the driving data, sensor data and image information before and after the escape; storing in the local storage unit according to the time stamp; used for subsequent AI model training, debugging and remote diagnosis analysis.

[0087] Further, after the escape is successful, a strategy for restoring the normal control mode is also included, which is specifically: by judging that Sk returns to below the normal driving threshold for more than T_rec; gradually releasing the AI takeover right of the throttle and direction control, and returning to the manual or conventional control mode; recording the escape control log this time, and clearing the temporary cache data.

[0088] In summary, the embodiment collects driving data in real time through multi-modal sensors and calculates skid determination indicators, and dynamically generates adaptive escape strategies combined with three-dimensional road surface information, which solves the problems of inaccurate skid determination and lagging control strategy of traditional methods, and has the advantages of improving the escape efficiency and safety.

[0089] Reference Figure 3 The application further provides a vehicle skid escape control device, which includes: The collection module 310 is configured to collect current driving data of the vehicle through a plurality of multi-modal sensors arranged on the vehicle, and the driving data includes the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel rotation angle, the body posture parameter, and the GPS displacement change value. The calculation module 320 is configured to calculate a slip determination index of each tire based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel rotation angle, the body posture parameter, and the GPS displacement change value. The control module 330 is configured to control the vehicle to enter a slip escape mode after determining that the tire slips based on the slip determination index, collect road surface information in front of the vehicle and in a tire area, and generate a corresponding escape control strategy based on the road surface information and the slip determination index, and control the vehicle to drive to escape.

[0090] Optionally, the calculation module 320 includes: The first judgment unit 321 is configured to preliminarily determine whether the vehicle slips based on the body posture parameter and the GPS displacement change value. The calculation unit 322 is configured to calculate the slip determination index of each wheel based on the rotation speed of each wheel, the acceleration of the vehicle, and the steering wheel rotation angle by using a calculation formula of the slip determination index when it is determined that the vehicle preliminarily slips.

[0091] Optionally, the device further includes a second judgment unit 323 configured to: determine whether the slip determination index reaches a preset index value; if yes, calculate the wheel speed difference, the vehicle longitudinal acceleration change rate, and the body pitch angle, and construct a slip judgment function based on the wheel speed difference, the vehicle longitudinal acceleration change rate, and the body pitch angle; solve the slip judgment function to determine whether the vehicle is in a slip state; if yes, determine that the tire slips.

[0092] Optionally, the control module 330 includes an acquisition unit 331 configured to: acquire a three-dimensional terrain image in front of the vehicle and in a tire contact area through a laser radar, a camera, or an ultrasonic sensor; perform grid processing on the three-dimensional terrain image to extract road surface information of a region corresponding to each wheel, wherein the road surface information includes the current depression depth, the slope, and the road surface adhesion coefficient of the wheel position.

[0093] Optionally, the control module 330 further includes an escape unit 332 configured to: calculate a driving force distribution ratio of each wheel according to the slip determination index of each wheel and the corresponding slope and road surface adhesion coefficient; compare the road adhesion coefficient of each wheel with a preset driving force threshold value; based on the comparison result, the driving force distribution ratio of each wheel is optimized and adjusted, wherein the optimized and adjusted driving force distribution ratio includes reducing the driving force ratio of the wheel whose road adhesion coefficient is lower than the driving force threshold value and increasing the driving force ratio of the wheel whose road adhesion coefficient is higher than the driving force threshold value; dynamically adjust the throttle opening and steering wheel angle of the vehicle according to the optimized driving force ratio, and output the corresponding escape control strategy to control the vehicle to drive in the escape mode.

[0094] Optionally, the escape unit 332 is specifically used for: inputting the optimized driving force ratio, the current throttle opening of the vehicle, and the current depth of the wheel position, the road adhesion coefficient, and the vehicle attitude parameter into a combined model of a multilayer perception machine and a recurrent neural network to output a signal control sequence of throttle opening and steering wheel angle control; execute the signal control sequence to control the vehicle to drive in the escape mode.

[0095] Optionally, the device further includes an optimization module 340, configured to: monitor the vertical load of each wheel and the real-time depth of the wheel position in the escape process in real time; if the vertical load of a certain wheel is not greater than a set value and the real-time depth is greater than the current depth, the suspension height of the corresponding wheel is reduced to increase the ground contact area, and the vehicle continues to drive in the escape mode; if the vertical load of a certain wheel is greater than a set value, the suspension height of the corresponding wheel is increased until the vehicle cabin is in a horizontal position balance state, and the vehicle continues to drive in the escape mode.

[0096] Through the implementation of the above technical solutions, the current driving data of the vehicle is collected by the multi-modal sensor arranged on the vehicle, the driving data including the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameter, and the GPS displacement change value; the slip determination index of each tire is calculated based on the rotation speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameter, and the GPS displacement change value; after determining the tire slip based on the slip determination index, the vehicle enters the slip escape mode, and the road surface information in front of the vehicle and the tire area is collected; the corresponding escape control strategy is generated based on the road surface information and the slip determination index, and the vehicle is controlled to drive in the escape mode. The application collects the driving state of the vehicle and the road surface information in real time through the multi-modal sensor, combines the slip determination index calculation formula and the artificial intelligence dynamic control strategy, realizes real-time detection of the slip state, and ensures the real-time performance of the escape control and the success rate of the vehicle escape.

[0097] The above Figure 4The vehicle slip escape control device in the present application is described in detail from the perspective of the modular functional entity, and the vehicle slip escape control device in the present application is described in detail from the perspective of hardware processing.

[0098] Referring to Figure 4 The vehicle slip escape control device shown in the figure includes a processor 400 and a memory 401, the memory 401 stores machine executable instructions that can be executed by the processor 400, and the processor 400 executes the machine executable instructions to realize the vehicle slip escape control method described above.

[0099] Further, Figure 4 The vehicle slip escape control device shown in the figure further includes a bus 402 and a communication interface 403, and the processor 400, the communication interface 403 and the memory 401 are connected through the bus 402.

[0100] The memory 401 can include a high-speed random access memory (RAM), and can also include a non-volatile memory, for example, at least one disk memory. The communication connection between the system node and at least one other node is realized through at least one communication interface 403 (which can be wired or wireless), and the Internet, a wide area network, a local network, a metropolitan area network, etc. can be used. The bus 402 can be an ISA bus, a PCI bus or an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0101] The processor 400 can be an integrated circuit chip having a processing capability of signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit of hardware in the processor 400 or the instruction in the form of software. The processor 400 described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. Each method, step and logic block diagram disclosed in the embodiments of the present disclosure can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with the embodiments of the present disclosure can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read only memory, a programmable read only memory or an electrically erasable programmable memory, a register, etc. The storage medium in the art. The storage medium is located in the memory 401, and the processor 400 reads the information in the memory 401, and combines the hardware to complete the method steps of the foregoing embodiments.

[0102] The present application also provides a computer readable storage medium, which can be a non-volatile computer readable storage medium, and can also be a volatile computer readable storage medium, and the computer readable storage medium has instructions stored therein, and when the instructions are run on a computer, the computer executes the steps of the vehicle slipping escape control method.

[0103] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above described system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0104] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0105] The above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for controlling vehicle skidding and getting out of trouble, characterized in that, The method includes: The vehicle's current driving data is collected by multimodal sensors installed on the vehicle. The driving data includes the rotational speed of each wheel, the vehicle's acceleration, the steering wheel angle, the vehicle's attitude parameters, and the GPS displacement change value. The slippage determination index of each tire is calculated based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body posture parameters, and the GPS displacement change value. After determining tire slippage based on the slippage determination index, the vehicle is controlled to enter the slippage escape mode, and road surface information in front of the vehicle and the tire area is collected. Based on the road surface information and the slippage determination index, a corresponding get-out-of-trouble control strategy is generated, and the vehicle is controlled to perform get-out-of-trouble driving.

2. The vehicle skidding and escaping control method according to claim 1, characterized in that, The calculation of slippage judgment indicators for each tire based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameters, and the GPS displacement change value includes: Based on the vehicle posture parameters and the GPS displacement change value, a preliminary judgment is made as to whether the vehicle is skidding. If so, the slippage determination index of each wheel is calculated based on the rotational speed of each wheel, the acceleration of the vehicle, and the steering wheel angle using the calculation formula of the slippage determination index.

3. The vehicle skidding and escaping control method according to claim 1, characterized in that, After calculating the slippage determination index for each tire based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameters, and the GPS displacement change value, the method further includes: Determine whether the slippage determination index has reached the preset index value; If the condition is met, the wheel speed difference, the rate of change of vehicle longitudinal acceleration, and the vehicle pitch angle are calculated, and a slip judgment function is constructed based on the wheel speed difference, the rate of change of vehicle longitudinal acceleration, and the vehicle pitch angle. Solve the slip judgment function to determine whether the vehicle is in a slipping state; If so, then the tires are definitely slipping.

4. The vehicle skidding and escaping control method according to claim 1, characterized in that, The collection of road surface information in front of the vehicle and in the tire area includes: Three-dimensional terrain images of the area in front of the vehicle and the tire contact area are obtained using lidar, cameras, or ultrasonic sensors. The three-dimensional terrain image is processed into a grid to extract road surface information for each wheel's corresponding area. The road surface information includes the current depression depth, slope, and road surface adhesion coefficient at the wheel's position.

5. The vehicle skidding and escaping control method according to claim 4, characterized in that, The step of generating a corresponding traction control strategy based on the road surface information and the slippage determination index, and controlling the vehicle to perform traction control driving, includes: Based on the slippage judgment index of each wheel and the corresponding slope and road adhesion coefficient, calculate the driving force distribution ratio of each wheel. The road adhesion coefficient of each wheel is compared with a pre-set driving force threshold. Based on the comparison results, the driving force distribution ratio of each wheel is optimized and adjusted. The optimization and adjustment include reducing the driving force ratio of wheels with road surface adhesion coefficient below the driving force threshold and increasing the driving force ratio of wheels with road surface adhesion coefficient above the driving force threshold. The vehicle's throttle opening and steering wheel angle are dynamically adjusted based on the optimized driving force ratio, and corresponding traction control strategies are output to control the vehicle to extricate itself from difficult situations.

6. The vehicle skidding and escaping control method according to claim 5, characterized in that, The process of dynamically adjusting the vehicle's throttle opening and steering wheel angle based on the optimized driving force ratio, and outputting a corresponding traction control strategy to control the vehicle for traction control, includes: The optimized driving force ratio, the vehicle's current throttle opening, the current indentation depth of the wheel position, the road adhesion coefficient, and the vehicle attitude parameters are input into a pre-trained multilayer perceptron and recurrent neural network combination model to output a signal control sequence for throttle opening and steering wheel angle control. The signal control sequence is executed to control the vehicle to extricate itself from a difficult situation.

7. The vehicle skidding and escaping control method according to claim 6, characterized in that, After executing the signal control sequence to control the vehicle for extrication driving, the method further includes: Real-time monitoring of the vertical load on each wheel and the real-time indentation depth at the wheel position during the escape process; If the vertical load on a wheel is not greater than the set value and the real-time dent depth is greater than the current dent level, the suspension height of the corresponding wheel will be reduced to increase the contact area, and the vehicle will continue to extricate itself from the ditch. If the vertical load on a certain wheel exceeds the set value, the suspension height of the corresponding wheel will be increased until the vehicle body is in a level and balanced state, and the vehicle will continue to extricate itself from the predicament.

8. A vehicle skidding and escaping control device, characterized in that, The vehicle skidding and getting out of trouble control device includes: The data acquisition module is used to collect the vehicle's current driving data through multimodal sensors installed on the vehicle. The driving data includes the rotational speed of each wheel, the vehicle's acceleration, the steering wheel angle, the vehicle's attitude parameters, and the GPS displacement change value. The calculation module is used to calculate the slippage judgment index of each tire based on the rotational speed of each wheel, the acceleration of the vehicle, the steering wheel angle, the vehicle body attitude parameters and the GPS displacement change value. The control module is used to control the vehicle to enter the skid escape mode after determining that the tires are skidding based on the skid judgment index, and to collect road surface information in front of the vehicle and the tire area; and to generate a corresponding escape control strategy based on the road surface information and the skid judgment index, and to control the vehicle to perform escape driving.

9. A vehicle skidding and escaping control device, characterized in that, The vehicle skidding and getting out of trouble control device includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the vehicle skidding and escaping control device to execute the vehicle skidding and escaping control method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instruction is read and executed, it performs the vehicle skidding and getting out of trouble control method as described in any one of claims 1-7.

Images