Vehicle suspension control method and device, electronic equipment and storage medium

Abstract

The invention relates to a vehicle suspension control method and device, electronic equipment and a storage medium. The vehicle suspension control method comprises the following steps: acquiring obstacle information of a preview obstacle and an obstacle identification confidence coefficient, which are obtained by performing preview identification on a vehicle driving road surface; the optimal control force of an actuator of a suspension in the vehicle is determined according to the obstacle information; determining a target confidence interval to which the obstacle recognition confidence belongs in a plurality of preset confidence intervals; and according to the target confidence interval and the optimal control force, determining the actual control force of the suspension. According to the embodiment of the invention, hierarchical control of the actual output force of the suspension can be carried out according to different obstacle recognition confidence coefficients, so that control of the actual output force of the suspension is matched with the obstacle recognition confidence coefficients, the control of the actual output force of the suspension better meets the actual driving working condition requirements of the vehicle, the vehicle comfort is improved, and the vehicle safety is improved. The smoothness of the vehicle passing through the obstacle is improved, and the energy consumption of suspension control is saved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle active suspension control, and particularly relates to a vehicle suspension control method, device, electronic device, and storage medium. Background Art

[0002] When a vehicle is driving on an urban paved road surface, single-impact road surfaces such as speed bumps, road surface bumps and depressions, cracks, and road surface repairs will have a great impact on comfort. The electronically controlled suspension adjusts the suspension force in real time according to the vehicle dynamics information input by the sensor to relieve the road surface impact and reduce the body jolt. Since this is a control action based on the vehicle information that has been changed by the road surface input, its control has hysteresis, and due to the sensor information transmission, controller calculation, and actuator action response delay, this hysteresis is amplified, resulting in a poor control effect and the deterioration of the vehicle ride comfort. Accordingly, in order to improve this hysteresis, feedforward control is applied to the suspension. The active suspension with visual preview detects the road surface information ahead through the visual preview system and inputs it to the suspension controller, and the suspension controller adjusts the suspension according to the road surface information ahead.

[0003] However, the traditional preview suspension control system may fail to accurately identify obstacles. When misidentifying existing obstacles, there is no control action on the possible pulse road surface, and the due role and performance of the preview control system cannot be exerted. Or, when misidentifying non-existent obstacles, the full active suspension actuation will generate reverse excitation, resulting in the deterioration of comfort, and it is impossible to balance the misidentification risk and the ride comfort requirement. Summary of the Invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present application provides a vehicle suspension control method, device, electronic device, and storage medium.

[0005] In a first aspect, the present application provides a vehicle suspension control method, including: Obtaining the obstacle information and obstacle recognition confidence of the preview obstacle obtained by pre-scanning and identifying the road surface on which the vehicle travels; Determining the optimal control force of the actuator of the suspension in the vehicle according to the obstacle information; Determining the target confidence interval to which the obstacle recognition confidence belongs among a plurality of pre-set confidence intervals; Determining the actual control force of the suspension according to the target confidence interval and the optimal control force.

[0006] Optionally, the obstacle information includes: the longitudinal distance of the obstacle, the vertical height of the obstacle, the longitudinal width of the obstacle, and the time correction parameter. Determining the optimal control force of the actuator of the suspension in the vehicle according to the obstacle information includes: Determine the road surface elevation curve based on the longitudinal distance of the obstacle, the vertical height of the obstacle, and the longitudinal width of the obstacle; Convert the road surface elevation curve from the spatial domain to the road surface excitation information in the time domain based on the current time correction parameter; Input the road surface excitation information into the 1 / 4 model of the vehicle suspension system to obtain the suspension dynamic performance parameters; Determine the optimal control force of the actuator according to the suspension dynamic performance parameters and the preset optimal control algorithm.

[0007] Optionally, determining the actual control force for the suspension according to the target confidence interval and the optimal control force includes: Determine a first suspension control mode for the suspension according to the target confidence interval and activate the first suspension control mode; Obtain the current suspension speed of the suspension, and determine the current maximum damping force and minimum damping force of the shock absorber of the suspension according to the current suspension speed; When the first suspension control mode is a non-silent control mode, determine the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force.

[0008] Optionally, if the target confidence interval is a first confidence interval and the first suspension control mode is a fully active suspension control mode, determining the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, control the shock absorber of the suspension to output the damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, control the shock absorber of the suspension to output the maximum damping force, and control the vehicle front wheel actuator to output the actuation force according to the difference between the optimal control force and the maximum damping force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, control the shock absorber of the suspension to output the minimum damping force, and control the vehicle front wheel actuator to output the actuation force according to the sum of the optimal control force and the maximum damping force.

[0009] Optionally, if the target confidence interval is a second confidence interval, the maximum boundary of the second confidence interval is less than the minimum boundary of the first confidence interval, and the first suspension control mode is a semi-active suspension control mode; determining the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, control the shock absorber of the suspension to output the damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, control the shock absorber of the suspension to output the maximum damping force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, control the shock absorber of the suspension to output the minimum damping force.

[0010] Optionally, if the target confidence interval is the third confidence interval and the maximum boundary of the third confidence interval is less than the minimum boundary of the second confidence interval, the first suspension control mode is a low-gain semi-active suspension control mode; determining the actual control force of the suspension based on the optimal control force, the maximum damping force and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the optimal control force and a preset confidence coefficient to obtain a first output force, and control the shock absorber of the suspension to output the damping force according to the first output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the maximum damping force and a preset confidence coefficient to obtain a second output force, and control the shock absorber of the suspension to output the damping force according to the second output force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, calculate the product of the minimum damping force and a preset confidence coefficient to obtain a third output force, and control the shock absorber of the suspension to output the damping force according to the third output force.

[0011] Optionally, if the target confidence interval is the fourth confidence interval and the maximum boundary of the fourth confidence interval is less than the minimum boundary of the third confidence interval, the first suspension control mode is a high-gain semi-active suspension control mode; determining the actual control force of the suspension of the vehicle based on the optimal control force, the maximum damping force and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the square of the preset confidence coefficient and the optimal control force to obtain a fourth output force, and control the shock absorber of the suspension to output the damping force according to the fourth output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the square of the preset confidence coefficient and the maximum damping force to obtain a fifth output force, and control the shock absorber of the suspension to output the damping force according to the fifth output force; Alternatively, when the optimal control force is opposite to the damping force of the suspension, calculate the product of the square of the preset confidence coefficient and the minimum damping force to obtain a sixth output force, and control the shock absorber of the suspension to output damping force according to the sixth output force.

[0012] Optionally, the method further includes: Obtain a plurality of first unsprung accelerations of the front lower control arm within a first time window; Determine whether the vehicle has traveled on a pulse road surface including pulse obstacles according to the plurality of first unsprung accelerations; If it is determined that the vehicle has traveled on a pulse road surface including pulse obstacles, perform spatio-temporal consistency matching on the preview obstacle and the pulse obstacle according to the plurality of first unsprung accelerations within the first time window and the obstacle information; If the spatio-temporal consistency matching between the preview obstacle and the pulse obstacle is successful, update the time correction parameter according to the plurality of first unsprung accelerations within the first time window, and update the obstacle recognition confidence according to a preset update rule.

[0013] Optionally, determining whether the vehicle has traveled on a pulse road surface including pulse obstacles according to the plurality of first unsprung accelerations includes: Extract the first unsprung accelerations located within a second time window from the plurality of first unsprung accelerations within the first time window, and the second time window is located within the first time window; Determine a first acceleration steady-state value within a long time window according to the first unsprung accelerations within the first time window; Determine a second acceleration steady-state value within a short time window according to the first unsprung accelerations within the second time window; Determine whether the ratio of the first acceleration steady-state value to the second acceleration steady-state value is greater than or equal to a preset steady-state threshold; If the ratio is greater than or equal to the preset threshold, determine that the front wheel of the vehicle has traveled on a pulse road surface.

[0014] Optionally, performing spatio-temporal consistency matching on the preview obstacle and the pulse obstacle according to the plurality of first unsprung accelerations within the first time window and the obstacle information includes: Determine the pulse edge time when the wheel travels onto the pulse obstacle according to the plurality of first unsprung accelerations within the first time window; If the pulse edge time is within a preset time range determined according to the obstacle information and the dynamic error of the longitudinal distance of the obstacle conforms to a normal distribution, determine a first concave-convex attribute of the preview obstacle according to the obstacle information, and determine a second concave-convex attribute of the pulse obstacle according to the plurality of first unsprung accelerations within the first time window; Determine whether the first concavo-convex attribute is consistent with the second concavo-convex attribute; If the first concavo-convex attribute is consistent with the second concavo-convex attribute, determine that the spatio-temporal consistency matching between the preview obstacle and the pulse obstacle passes.

[0015] Optionally, determining the first concavo-convex attribute of the preview obstacle according to the obstacle information, and determining the second concavo-convex attribute of the pulse obstacle according to multiple first unsprung accelerations within the first time window, includes: Determine the first concavo-convex attribute of the preview obstacle according to the obstacle type in the obstacle information; Calculate the acceleration mean value of the first unsprung acceleration within the second time window in the first time window; Determine the second concavo-convex attribute of the pulse obstacle according to the positive or negative of the acceleration mean value.

[0016] Optionally, updating the time correction parameter according to multiple first unsprung accelerations within the first time window, includes: Determine the first moment when the wheel travels onto the pulse obstacle according to multiple first unsprung accelerations within the first time window; Estimate the second moment when the wheel travels onto the preview obstacle according to the obstacle information; Calculate the time difference between the first moment and the second moment, and update the time correction parameter based on the time difference; Determine the confidence range corresponding to the obstacle recognition confidence, and update the obstacle recognition confidence according to the update rule corresponding to the confidence range.

[0017] In a second aspect, the present application provides a vehicle suspension control device, including: An acquisition module, configured to acquire obstacle information and obstacle recognition confidence of a preview obstacle obtained by pre-scanning and recognizing a road surface on which a vehicle travels; A first determination module, configured to determine an optimal control force of an actuator of a suspension in the vehicle according to the obstacle information; A second determination module, configured to determine a target confidence interval to which the obstacle recognition confidence belongs among multiple pre-set confidence intervals; A third determination module, configured to determine an actual control force on the suspension according to the target confidence interval and the optimal control force.

[0018] In a third aspect, the present application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used for storing a computer program; A processor, when executing a program stored in a memory, implements the vehicle suspension control method according to any one of the first aspects.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, on which a program of a vehicle suspension control method is stored. When the program of the vehicle suspension control method is executed by a processor, the steps of the vehicle suspension control method according to any one of the first aspects are implemented.

[0020] Advantages of the present invention: In the embodiment of the present application, by using the obstacle information pre-identified from the road surface on which the vehicle travels to determine the optimal control force of the actuator of the suspension, and using the obstacle recognition confidence obtained by pre-identification to determine the target confidence interval among multiple pre-set confidence intervals, and then determining the actual control force of the suspension according to the target confidence interval and the optimal control force. By setting multiple candidate pre-set confidence intervals, hierarchical control of the actual output force of the suspension is realized according to different obstacle recognition confidences, so that the control of the actual output force of the suspension matches the obstacle recognition confidence, making the control of the actual output force of the suspension more in line with the actual driving conditions of the vehicle, improving the vehicle comfort, improving the smoothness of the vehicle when passing obstacles, and saving the energy consumption of suspension control. Description of the Drawings

[0021] The drawings here are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present invention, and are used together with the specification to explain the principles of the present invention.

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0023] Figure 1 It is a flowchart of a vehicle suspension control method provided by an embodiment of the present application; Figure 2 It is a schematic diagram of determining a road surface elevation curve according to obstacle information provided by an embodiment of the present application; Figure 3 It is a schematic diagram of a 1 / 4 model of a vehicle suspension system provided by an embodiment of the present application; Figure 4 It is a schematic diagram of multi-level threshold arbitration decision for obstacle recognition confidence provided by an embodiment of the present application; Figure 5 It is a flowchart of another vehicle suspension control method provided by an embodiment of the present application; Figure 6 Structural diagram of a vehicle suspension control device provided by an embodiment of the present application; Figure 7 Structural diagram of an electronic device provided by an embodiment of the present application. Specific implementation manners

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some but not all of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0025] Since the traditional preview suspension control system may fail to accurately identify obstacles. When misidentifying existing obstacles, there is no control action for possible pulse roads, and the due functions and performance of the preview control system cannot be exerted. Or, when misidentifying non-existing obstacles, full-active suspension actuation will generate reverse excitation, resulting in deteriorated comfort, and it is impossible to balance the misidentification risk and the ride comfort requirement. For this reason, the embodiments of the present application provide a vehicle suspension control method, device, electronic device, and storage medium.

[0026] The embodiments of the present application provide a vehicle suspension control method, as Figure 1 shown, including: Step S101, obtaining obstacle information and obstacle recognition confidence of preview obstacles obtained by pre-viewing and recognizing the road surface on which the vehicle travels; In the embodiments of the present application, the forward preview device is installed at a position with a good view in the front of the vehicle, such as the glass in front of the rearview mirror, above the bumper, etc. The forward preview device may exemplarily include sensors such as cameras, radars, lidars, and machine learning models (such as CNN, RNN). The forward preview device is used to obtain pulse road surface information on the tire travel trajectory in the vehicle travel direction, use the machine learning model to recognize road surface features (such as image texture, point cloud height difference), generate a confidence score of the obstacle type while outputting the recognition result (for example, the CNN model recognizes a road pothole, outputs a "pothole" label and the corresponding confidence of the obstacle type), mark the recognized pulse road surface as an obstacle, number the obstacle and continuously track the obstacle until the vehicle's rear axle drives away from the obstacle by a certain distance and then releases the obstacle information, and the preview recognition obtains the obstacle information and the obstacle recognition confidence , taking the preview information on the left wheel trajectory of the vehicle as an example, the obstacle information includes: obstacle number i , obstacle type , longitudinal distance of the obstacle , the vertical height of the obstacle and the longitudinal width of the obstacle etc.

[0027] Among them, the obstacle number i , the forward preview device can only identify and mark a limited number of obstacles at the same time. When the marking limit N is exceeded, the information of the marked obstacles will be released before marking again, and so on. When the number overflows, the numbering will start from 0 again.

[0028] Obstacle type , such as pulse roads like speed bumps, protrusions, potholes, damages, etc. The obstacle type is represented by natural numbers etc. Among them, 0 represents a general random road surface. Raised road surface types such as speed bumps, protrusions, and raised manhole covers are represented by even numbers, and sunken road surface types such as potholes, damaged road surfaces, and sunken manhole covers are represented by odd numbers.

[0029] Obstacle longitudinal distance , the preview device calculates the longitudinal distance between the obstacle and the rear axle of the vehicle by fusing the information of the vision camera or the vision camera combined with sensors such as millimeter-wave radar and lidar.

[0030] The vertical height of the obstacle and the longitudinal width , the preview device calculates the height information of the uneven road surface by fusing the information of sensors such as vision camera, millimeter-wave radar, and lidar.

[0031] Confidence level of obstacle event , the credibility of detecting and identifying abnormal unevenness of the road surface, represented by 0~1.

[0032] In addition, the main information that can be obtained through sensors, CAN communication or calculation processing also includes: unsprung acceleration information, suspension height, suspension speed, vehicle speed information, for use in subsequent embodiments.

[0033] Step S102, determine the optimal control force of the actuator of the suspension in the vehicle according to the obstacle information; In this step, the road elevation curve can be determined according to the obstacle longitudinal distance, the obstacle vertical height and the obstacle longitudinal width; convert the road elevation curve from the spatial domain to the time-domain road excitation information based on the current time correction parameter; input the road excitation information into the 1 / 4 model of the vehicle suspension system to obtain the suspension dynamic performance parameters; determine the optimal control force of the actuator (including the actuator and the shock absorber) according to the suspension dynamic performance parameters and the preset optimal control algorithm.

[0034] Further, the suspension controller can calculate an approximate road surface elevation curve based on the longitudinal distance of the obstacle, the vertical height of the obstacle, and the longitudinal width of the obstacle identified by preview recognition. Here, the road surface elevation curve is the information of the vertical height varying with the longitudinal distance on the driving trajectory of the vehicle tires in the vehicle driving direction. Then, according to the preset time delay parameter and the current time correction parameter, the road surface elevation information is converted from the spatial domain description information into the road surface excitation information in the time domain for the wheels. The road surface excitation information of the left and right wheels in the time domain is input into the suspension control system, and the suspension control system inputs the road surface excitation information into the 1 / 4 model of the vehicle suspension system to obtain the suspension dynamic performance parameters. The suspension dynamic performance parameters include: the vehicle body acceleration representing ride comfort, the wheel dynamic deflection representing handling stability, and the suspension dynamic deflection representing driving safety. The weighted calculation of the suspension dynamic performance parameters is performed through the optimal control algorithm, i.e., the LQR control algorithm. The integral value of the weighted sum of squares of the above three suspension dynamic performance parameters in the time domain is used as the target performance index, and the optimal suspension force input is obtained by solving.

[0035] In practical applications, first, the input displacement information of the pulsed road surface ahead can be calculated based on the obstacle type, distance, height, and width information. The preview distance in the preview information is the distance from the rear wheel to the obstacle. Therefore, according to the preview distance and the vehicle wheelbase the distance information from the front wheel to the obstacle is obtained, and further the road surface excitation input function of the front wheel is obtained .

[0036] Pulsed road surfaces such as road surface bumps, speed bumps, and potholes are all sinusoidal-like fluctuations and are described by a cosine function because its slope is zero near the starting and ending points, as shown in Scenario 1 in Figure 2 : (1) i Obstacle number ， Obstacle type, Longitudinal distance of the obstacle, Vertical height of the obstacle, Longitudinal width, Vehicle wheelbase Pulsed road surfaces such as raised manhole covers, sunken manhole covers, and ground cracks have a trapezoidal cross-section, and the slope near their starting and ending points changes suddenly, and the middle is relatively flat, as shown in Scenario 2 in Figure 2 . For simplicity of description, a step function is used to describe it: (2) Use the road surface elevation input related to the preview distance input as the front suspension control input. On the one hand, consider the measured comprehensive delay of the information reception program operation and actuator action , it should enter the preview control state in advance when the front wheel approaches the pulse road surface. On the other hand, considering the inevitable errors in other aspects such as information transmission delay, there is a systematic error (i.e., the time correction parameter) between the time to reach the obstacle estimated based on the preview distance and the actual time to reach the obstacle. , this time correction parameter can obtain its initial value by calibration and be corrected in the subsequent embodiments.

[0037] Considering the delay and the system error , the distance determination formula for the front wheel to enter the preview control state is obtained: (3) where is the longitudinal acceleration.

[0038] Make corrections to the preview distance information (4) Substitute the corrected preview distance into Equation (1) and Equation (2), and the road surface excitation input for the front wheel preview control is obtained .

[0039] After obtaining the road surface excitation input, the suspension system control force is obtained according to the 1 / 4 model of the vehicle suspension system and the LQR control algorithm.

[0040] Among them, the 1 / 4 model of the vehicle suspension system is as Figure 3 shown, Figure 3 in and are the stiffness respectively, is the passive damping coefficient of the suspension system, is the sprung mass, is the unsprung mass, is the sprung mass displacement, is the unsprung mass displacement, is the road surface excitation displacement. Establish the dynamic differential equation of the suspension system: (5) (6) Take the state vector , input the pulse road surface excitation into the vehicle suspension model to obtain the output variable

[0041] Convert the differential equation of the vehicle suspension system into a system state equation and an output equation: (7) (8) In the formula: ; ; ; ; ; ;

[0042] Among the output variables Y, the three variables are respectively the car body acceleration representing ride comfort, the wheel dynamic deflection representing handling stability, and the suspension dynamic deflection representing driving safety. Using the LQR control algorithm, the integral value of the weighted sum of squares of the above three performance evaluation indexes in the time domain is used as the target performance index, which is expressed as follows: (9) In the formula, is the initial control time; are respectively the weighted coefficients of the car body acceleration, the wheel dynamic deflection and the suspension dynamic deflection.

[0043] Substituting formula (8) into formula (9), the performance index can be transformed into: (10) In the formula: ; ; ; .

[0044] From formula (10), the optimal control matrix of the actuator can be obtained as: (11) In the formula, K is the optimal state feedback gain matrix; P is the solution of the following form of Riccati equation: (12) Using the LQR control algorithm to perform weighted calculation on the output variable Y, the optimal control force under the pulse road surface elevation input is obtained.

[0045] This application considers the hysteresis caused by various factors in the control system. When the distance to the obstacle approaches and breaks through the threshold, the input preview distance information is switched to the state predicted and updated by the vehicle longitudinal speed acceleration. This reduces the response hysteresis of the actuator to the input and improves the control system accuracy.

[0046] Step S103, determine the target confidence interval to which the obstacle recognition confidence belongs among multiple pre-set confidence intervals; In this step, the obstacle recognition confidence can be compared with the upper and lower boundaries of each preset confidence interval respectively, and the preset confidence interval in which the obstacle recognition confidence is greater than the lower boundary and less than the upper boundary is determined as the target confidence interval.

[0047] Multiple preset confidence intervals can be divided into five levels of confidence decision threshold determination: High confidence interval: Activate full active control (joint adjustment of height, stiffness + damping); Relatively high confidence interval: Activate semi-active control (only damping adjustment); Medium confidence interval: Activate conservative strategy semi-active control (only damping adjustment, and output damping force); Relatively low confidence interval: Activate energy-saving strategy semi-active control (only damping adjustment, and output damping force); Low confidence interval: The preview control module does not output an effective control signal amount.

[0048] Step S104, determine the actual control force for the suspension according to the target confidence interval and the optimal control force.

[0049] In this step, the first suspension control mode for the suspension can be determined according to the target confidence interval, and the first suspension control mode can be activated; obtain the current suspension speed of the suspension, and determine the current maximum damping force and minimum damping force of the shock absorber of the suspension according to the current suspension speed; when the first suspension control mode is a non-silent control mode, determine the actual control force of the suspension based on the optimal control force, the maximum damping force and the minimum damping force.

[0050] In the embodiments of the present application, the actual control force includes at least one of the driving force that the actuator actually needs to output and the damping force that the shock absorber needs to output.

[0051] For the suspension control at the front wheel position, a confidence-based hierarchical arbitration mechanism is adopted to decide the control method to be used. The suspension controller receives the information of obstacle i, and uses mean filtering for the confidence information to reduce its instability. The controller makes a judgment according to the current confidence, and decides to adopt the full active control method or the semi-active control method. When the confidence is in the high confidence interval, the suspension adopts the full active control method. When the confidence is in the relatively high confidence interval, the suspension adopts the semi-active control method. When the confidence is in the medium confidence interval, the suspension adopts the semi-active control method and multiplies the output result by the confidence. When the confidence is in the relatively low confidence interval, the suspension adopts the semi-active control method and multiplies the output result by the square of the confidence. When the confidence is in the low confidence interval, the preview control module of the suspension does not control the actuator.

[0052] The specific decision mechanism is as shown in the accompanying drawings of the specificationFigure 4 As shown, it includes a five-level decision threshold setting for decision-making and judgment through the arbitration module 1: In an implementation manner of the present application, if the target confidence interval is the first confidence interval (i.e., a high confidence interval, such as: ), and the first suspension control mode is a fully active suspension control mode, step S104 determines the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force, including: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, control the shock absorber of the suspension to output the damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, control the shock absorber of the suspension to output the maximum damping force, and control the front wheel actuator of the vehicle to output the actuation force according to the difference between the optimal control force and the maximum damping force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, control the shock absorber of the suspension to output the minimum damping force, and control the front wheel actuator of the vehicle to output the actuation force according to the sum of the optimal control force and the maximum damping force.

[0053] In practical applications, for the high confidence interval : Activate full active control (joint adjustment of actuation force and damping force), and switch to Figure 4 the execution mode of module 2 in According to the suspension speed, if the required suspension force is in the same direction as the damping force , then preferentially adjust the damping force to the maximum. The suspension actuation force only serves as compensation when the suspension damping force is insufficient to provide all the suspension force . This is because the energy consumption required to output the damping force is lower than that required to output the actuation force. The controller obtains the real-time maximum damping force based on the real-time suspension speed and the characteristic curve of the shock absorber. That is, if the required suspension force , then the damping force of the suspension shock absorber is sufficient to provide the required suspension force, the suspension actuator does not need to work, the actuation force , and the suspension damping force ; if , the suspension damping force , the damping force of the suspension shock absorber is insufficient to provide the required suspension force, and the suspension actuation force ; If is in the opposite direction to the damping force , then adjust the damping force to the shock absorber at the real-time suspension speed Lower minimum damping force That is , .

[0054] In an implementation manner of the present application, if the target confidence interval is the second confidence interval (i.e., the higher confidence interval, such as: ), the maximum boundary of the second confidence interval is less than the minimum boundary of the first confidence interval, and the first suspension control mode is a semi-active suspension control mode; step S104 determines the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force, including: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, control the shock absorber of the suspension to output the damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, control the shock absorber of the suspension to output the maximum damping force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, control the shock absorber of the suspension to output the minimum damping force.

[0055] In practical applications, for the higher confidence interval : Activate semi-active control (only damping adjustment, actuating force ), and switch to the execution mode of module 3 in Figure 4 ; In is in the same direction as the damping force , if , the damping force can provide all the required suspension force, then the suspension damping force , if , at this time the shock absorber cannot output enough damping force to meet the suspension force requirement, and can only adjust the damping force to the maximum ; In is in the opposite direction to the damping force , at this time cannot be achieved, then adjust the damping force to the lowest value , reducing the difference between the actual output force of the suspension and the required suspension force, that is .

[0056] In an implementation manner of the present application, if the target confidence interval is the third confidence interval (i.e., the medium confidence interval, such as: ), the maximum boundary of the third confidence interval is less than the minimum boundary of the second confidence interval, and the first suspension control mode is a low-damping semi-active suspension control mode; Step S104 determines the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force, including: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the optimal control force and a preset confidence coefficient to obtain a first output force, and control the shock absorber of the suspension to output damping force according to the first output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the maximum damping force and a preset confidence coefficient to obtain a second output force, and control the shock absorber of the suspension to output damping force according to the second output force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, calculate the product of the minimum damping force and a preset confidence coefficient to obtain a third output force, and control the shock absorber of the suspension to output damping force according to the third output force.

[0057] In practical applications, for the medium confidence interval, : Activate the conservative strategy semi-active control (only damping adjustment, actuation force ), and switch to Figure 4 the execution mode of module 4 in the middle.

[0058] And the output target damping force is the theoretical damping force multiplied by the confidence coefficient. Here, the theoretical damping force is the same as that in the case of a higher confidence level, then , because generally, the higher the damping of a variable damping shock absorber, the higher the energy consumption. This measure is to reduce the energy consumption of the suspension system.

[0059] In an embodiment of the present application, if the target confidence interval is the fourth confidence interval (i.e., the lower confidence interval, such as: ), the maximum boundary of the fourth confidence interval is less than the minimum boundary of the third confidence interval, and the first suspension control mode is a high-damping semi-active suspension control mode; Step S104 determines the actual control force of the vehicle suspension based on the optimal control force, the maximum damping force, and the minimum damping force, including: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the square of the preset confidence coefficient and the optimal control force to obtain the fourth output force, and control the shock absorber of the suspension to output the damping force according to the fourth output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the square of the preset confidence coefficient and the maximum damping force to obtain the fifth output force, and control the shock absorber of the suspension to output the damping force according to the fifth output force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, calculate the product of the square of the preset confidence coefficient and the minimum damping force to obtain the sixth output force, and control the shock absorber of the suspension to output the damping force according to the sixth output force.

[0060] In practical applications, for a lower confidence interval : Activate the energy-saving strategy semi-active control (only damping adjustment, actuator ), and switch to the execution mode of Module 5.

[0061] And the output target damping force is the theoretical damping force multiplied by the square of the confidence coefficient. Here, the theoretical damping force is the same as that in the case of a higher confidence level , then , and the purpose here is also to reduce energy consumption.

[0062] In an embodiment of the present application, if the target confidence interval is the fifth confidence interval (i.e., the low confidence interval ), the maximum boundary of the fifth confidence interval is less than the minimum boundary of the fourth confidence interval, and the first suspension control mode is the preview control silent mode; at this time, the preview control module does not output an effective control signal amount, and the actuator output force is based on the , issued by other control modules, and switches to Figure 4 the execution mode of Module 6 in it.

[0063] In summary, for the suspension control of the front wheel position, a confidence-based hierarchical arbitration mechanism is adopted to decide the control method to be used. The suspension controller receives the information of the obstacle. The controller needs to judge the confidence at this time and decide to adopt the full active control method or the semi-active control method. When the confidence is in the high confidence interval, the suspension adopts the full active control method. When the confidence is in the higher or medium confidence interval, the suspension adopts the semi-active control method and the lower the confidence, the lower the output force. When the confidence is in the low confidence interval, the preview control module of the suspension does not control the actuator. Based on the confidence-based hierarchical control strategy, the actuator output force in different confidence cases can be obtained and the damping force output by the shock absorber 。

[0064] In the embodiment of the present application, the arbitration of the suspension preview control mode is based on the confidence of image recognition of road obstacles. When the confidence is higher than the first confidence threshold, the preview suspension control system adopts the full active suspension control mode, and the full active suspension system's suppression effect on vehicle bumps is exerted by coordinately controlling the damping force and the actuating force of the suspension system. When the confidence is lower than the first confidence threshold and higher than the second confidence threshold, the preview suspension control system adopts the semi-active suspension control mode, and the smoothness of the vehicle passing through the obstacle is improved by controlling the damping of the suspension system to reduce the impact on the vehicle body when driving onto the obstacle and the aftershock after driving away from the obstacle. When the confidence is higher than the fourth confidence threshold and lower than the second confidence threshold, the preview suspension control system still adopts the semi-active suspension control mode, but multiplies the finally output damping force by a correlation coefficient related to the confidence to reduce the damping force output by the control algorithm. And the associated damping coefficient is switched with the third confidence threshold as the boundary. When it is lower than the fourth confidence threshold, no effective control signal quantity is output. In the decision-making of the suspension preview control mode of the vehicle front axle, the confidence of preview recognition is the main basis.

[0065] Due to the conventional preview control method, if the recognition confidence is lower than a certain threshold, it is judged to be unreliable, and the obstacle data will not be issued. However, there is a decrease in the recognition confidence of obstacles in the same scene due to changes in external light or environment. If there is an obstacle in front and the data is discarded because the recognition confidence is not high, the missed recognition does not trigger the preview control of the vehicle, which reduces the comfort. Therefore, the purpose of this application is to optimize the problem that relying on a single recognition confidence in the preview control of the active suspension cannot balance the comfort and the risk of misidentification well. Different control modes are adopted under multi-level preview recognition confidence thresholds. Since the fully active suspension actuator will quickly and significantly change the suspension height when actuated, if the flat ground is mistakenly identified as an obstacle, it will cause the vehicle to produce unnecessary shaking. Therefore, the fully active suspension control mode is adopted under high confidence conditions to give full play to the performance of the fully active suspension, and to avoid triggering the active action of the fully active suspension when misidentified as much as possible, and to minimize the impact of the vehicle on the pulse road surface to improve comfort. When the confidence is not very high, the semi-active suspension control mode is used. In the semi-active suspension control mode, only the damping force of the suspension system is adjusted. The damping force of the suspension system is a passive force, which generates a reverse force according to the relative movement speed of the spring and the unsprung. The advantage of this is that if there is indeed a pulse road surface, that is, there is an obstacle at this time, adjusting the damping can reduce the impact force and suppress the shaking of the vehicle body. If the recognition error does not actually exist in the impact road surface or the information obtained by the recognition is inaccurate, that is, when the confidence is not high, there is indeed no obstacle, and the road surface is flat. The damping force is the force opposite to the direction of suspension movement. It only dissipates energy and does not input additional energy. Adjusting the damping will not have any effect and will not produce a reverse excitation effect. When the recognition is medium or low, the optimal theoretical damping force calculated by the algorithm is passed through the reduction link, and when the recognition is low, the preview control is directly not outputting the effective control amount. Its role is that the lower the recognition confidence, the lower the damping value of the suspension system, and the lower the current driving the shock absorber, so that the adjustable damping shock absorber can fully exert its ability to improve comfort and reduce energy loss.

[0066] The embodiment of the present application determines the optimal control force of the actuator of the suspension by utilizing the obstacle information obtained by previewing and identifying the road surface on which the vehicle is traveling, determines the target confidence interval in multiple preset confidence intervals by utilizing the obstacle recognition confidence obtained by previewing and identifying, and then determines the actual control force of the suspension according to the target confidence interval and the optimal control force. By setting multiple candidate preset confidence intervals, graded control of the actual output force of the suspension is achieved according to different obstacle recognition confidences, so that the control of the actual output force of the suspension is matched with the obstacle recognition confidence, so that the control of the actual output force of the suspension is more in line with the actual driving conditions of the vehicle, thereby improving the comfort of the vehicle, improving the smoothness of the vehicle when passing through obstacles, and saving energy consumption of suspension control.

[0067] Traditional vision preview systems (such as cameras and lidar) are vulnerable to environmental lighting, occlusion, or sensor noise when detecting the distance to a speed bump, resulting in relatively large distance measurement errors. The preview distance error will directly cause the suspension actuator to be triggered prematurely or late, unable to accurately cancel the road surface impact, resulting in an increase in the vertical acceleration of the vehicle body and a reduction in comfort. Simply relying on the visual sensor filtering algorithm or increasing sensor redundancy (such as multi-camera fusion) is costly and cannot dynamically compensate for real-time errors. Existing suspension control methods do not fully utilize the physical feedback signals (such as acceleration) when the wheels pass over a speed bump for closed-loop correction. Therefore, in one implementation of this application, as Figure 5 shown, the vehicle suspension control method further includes: Step S201, obtaining a plurality of first unsprung accelerations of the front wheel lower control arm within a first time window; Acceleration sensors are installed on the front wheel lower control arm, one at each of the left and right lower control arm positions to obtain the unsprung accelerations at the positions of the two wheels respectively and transmit them to the suspension controller.

[0068] Step S202, determining whether the vehicle has traveled over a pulse road surface containing pulse obstacles based on the plurality of first unsprung accelerations; In this step, the suspension controller determines whether it has passed over a pulse road surface by windowing and feature extraction of the front wheel unsprung vertical acceleration.

[0069] Further, the first unsprung accelerations located within a second time window can be extracted from the plurality of first unsprung accelerations within the first time window, the second time window being located within the first time window; determining a first acceleration steady-state value within a long time window based on the first unsprung accelerations within the first time window; determining a second acceleration steady-state value within a short time window based on the first unsprung accelerations within the second time window; determining whether the ratio of the first acceleration steady-state value to the second acceleration steady-state value is greater than or equal to a preset steady-state threshold; if the ratio is greater than or equal to the preset threshold, determining that the front wheels of the vehicle have traveled over a pulse road surface.

[0070] In practical applications, it is possible to determine whether a pulse road surface has been passed by detecting the front wheel unsprung vertical acceleration. The short-term energy input of a pulse road surface is much higher than that of a general random road surface, and the root mean square value RMS of the data can reflect the vibration energy level to a certain extent. If the RMS value of the acceleration data changes suddenly, it is determined that the wheel has traveled onto a pulse road surface. The controller records the unsprung vertical acceleration values at the past sampling moments and takes the past (6) to calculate the root mean square value of the acceleration at this moment through the sliding window weighted average method, calculating the past The iterative update formula for the root mean square value of acceleration at each sampling moment is as follows: (7) The mutation of the RMS value can be reflected by the ratio of the root mean square value within a short window period to the root mean square value within a long window period. P If the vehicle drives onto a pulsed road surface, then the RMS value within the short window time RMS will mutate compared to the steady state P value within the long window time. For the determination of a pulsed road surface based on an acceleration sensor, there is a determination formula for the mutation ratio of the root mean square of the pulsed edge acceleration (here, the short time window is defined as the past moments, and the long time window is defined as the past moments): (8) In this formula, is a constant threshold, m, n is also a constant, and its value is obtained through calibration.

[0071] If the ratio P is greater than or equal to the preset threshold , it is determined that the RMS value has mutated. The mutation of the RMS value also reflects the mutation of the energy of the forced vibration of the unsprung mass, indicating that the front wheel of the vehicle has driven over a pulsed road surface, that is, the wheel has driven onto an obstacle.

[0072] After the vehicle drives to the edge of the pulsed road surface, the unsprung mass will oscillate under the action of the tire elastic force and the suspension spring force. The unsprung mass acceleration value will oscillate at a frequency close to the natural frequency of the unsprung mass. And due to the damping effect of the suspension and the tire, the oscillation amplitude will decay at a relatively fixed ratio, and the acceleration extreme point will appear when the road surface excitation suddenly increases. For a bump, P the maximum value will appear at the first extreme point, while for a pothole, the maximum value will appear at the second extreme point. Record this maximum value . If within the subsequent T1 time , it is not recognized as a new pulsed road surface edge. The decay ratio K1 and the silent time T1 are both determined through calibration.

[0073] Step S203, if it is determined that the vehicle has driven over a pulsed road surface containing pulsed obstacles, perform spatio-temporal consistency matching on the preview obstacle and the pulsed obstacle according to the multiple first unsprung accelerations within the first window period and the obstacle information; In some embodiments of the present invention, when the acceleration detection identifies a pulsed road surface, the obstacle information identified by preview is compared with the pulsed road surface information obtained based on the unsprung acceleration. According to the preview distance information and the real-time vehicle speed, the theoretical range of the predicted front-wheel arrival time is estimated. If the pulsed edge event time identified by the acceleration detection is within the error range of the arrival time calculated based on the preview information, it is considered that the preview-identified obstacle event and the acceleration-identified pulsed road surface edge event have the same obstacle event characteristics.

[0074] Furthermore, the concave-convex attributes of the obstacle type identified by preview are inspected against the concave-convex attributes determined by the unsprung acceleration detection. If the identifications are consistent, it is considered that the preview obstacle event and the acceleration-identified pulsed road surface edge event can be aligned, and this pulsed road surface edge event is marked as an alignable state.

[0075] If the pulsed edge event time identified by the acceleration detection is not within the expected time range of any preview obstacle event, or if the inspection of the concave-convex attributes of the obstacle type identified by preview against the concave-convex attributes determined by the unsprung acceleration detection results in an identification conflict, then this pulsed road surface edge event is marked as a non-alignable state, and this pulsed road surface edge event is not aligned with any preview obstacle event and is not used as a reference for error compensation and correction of the preview obstacle distance.

[0076] In this step, the pulsed edge moment when the wheel travels onto the pulsed obstacle can be determined based on multiple first unsprung accelerations within the first time window; if the pulsed edge moment is within the preset time range determined based on the obstacle information and the dynamic error of the longitudinal distance of the obstacle conforms to a normal distribution, the first concave-convex attribute of the preview obstacle is determined based on the obstacle information, and the second concave-convex attribute of the pulsed obstacle is determined based on multiple first unsprung accelerations within the first time window; it is determined whether the first concave-convex attribute and the second concave-convex attribute are consistent; if the first concave-convex attribute and the second concave-convex attribute are consistent, it is determined that the spatio-temporal consistency matching between the preview obstacle and the pulsed obstacle passes.

[0077] When the acceleration detection identifies a pulsed road surface, determine the longitudinal distance of the obstacle identified by preview of the dynamic error whether it conforms to a normal distribution , where the dynamic error of the preview-identified distance can be obtained in advance through experimental tests , and its 99% confidence interval of the dynamic error is obtained , and here can be regarded as the system steady-state error, and the 99% confidence interval of the preview distance is , that is is within the 99% confidence interval of the preview distance.

[0078] The preset time range can be determined according to the dynamic error range and the real-time vehicle speed. The theoretical range of the predicted front wheel arrival time, that is, the preset time range, is , if the pulse edge moment identified according to the acceleration detection is within the preset time range, that is .

[0079] Furthermore, the first convexity and concavity attribute of the preview obstacle can be determined according to the obstacle information, and the second convexity and concavity attribute of the pulse obstacle can be determined according to multiple first unsprung accelerations within the first time window, including: determining the first convexity and concavity attribute of the preview obstacle according to the obstacle type in the obstacle information; calculating the acceleration mean value of the first unsprung acceleration within the second time window in the first time window; determining the second convexity and concavity attribute of the pulse obstacle according to the positive or negative of the acceleration mean value, that is, judging whether the pulse road surface is a convex obstacle or a concave obstacle by the vertical acceleration direction of driving onto the edge of the pulse road surface.

[0080] Calculate the mean value of the unsprung acceleration at the past m moments: (9) If , the pulse road surface is identified as a convex road surface; If , the pulse road surface is identified as a concave road surface; If , then take the mean value of the unsprung acceleration at the past m-1 moments for judgment, where l is a calibrated value and l is less than m.

[0081] In summary, the suspension controller can judge whether the pulse road surface is a convex obstacle or a concave obstacle by the vertical acceleration direction of driving onto the edge of the pulse road surface.

[0082] When determining whether the first convexity and concavity attribute and the second convexity and concavity attribute are consistent, the convexity and concavity attribute of the preview-identified obstacle type can be tested with the convexity and concavity attribute judged by the unsprung acceleration detection, that is: (10) If the recognition is consistent, it is considered that the preview obstacle event and the acceleration recognition pulse road surface edge event can be aligned, and this pulse road surface edge event is marked as an alignable state. It should be noted that the purpose of the spatio-temporal consistency alignment is to use the preview information for the rear wheel control and the next front wheel control for correction.

[0083] If it is the actual pulse edge event time identified according to the acceleration detection , that is, if the pulsed road surface edge event does not occur within the expected event range of any preview obstacle event, or an identification conflict result is obtained by verifying the convexity and concavity attributes of the preview-identified obstacle type and the convexity and concavity attributes determined by the under-spring acceleration detection, then this pulsed road surface edge event is marked as in a non-alignable state, and this pulsed road surface edge event is not aligned with any preview obstacle event and is not used as a reference for error compensation and correction of the preview obstacle distance.

[0084] Step S204, if the spatio-temporal consistency between the preview obstacle and the pulsed obstacle is successfully matched, update the time correction parameter according to multiple first under-spring accelerations within the first time window, and update the obstacle identification confidence level according to a preset update rule.

[0085] In the embodiment of the present application, the actual time difference between the preview obstacle event and the pulsed obstacle event in the event can be calculated, and this difference is used as the correction value for the time when the rear wheel reaches the obstacle, and the system error of estimating the time to reach the obstacle according to the preview distance is updated with this difference data. When in an alignable state and the convexity and concavity attributes are identified consistently, it can be considered that the preview-identified obstacle information has a high credibility, and the confidence level of the preview obstacle information will be updated. For the rear wheel passing the obstacle i the same confidence-based hierarchical control strategy as that of the front wheel is adopted.

[0086] In this step, the first moment when the wheel travels to the pulsed obstacle can be determined according to multiple first under-spring accelerations within the first time window. That is to say, when the wheel travels to the pulsed road surface, the moment when the threshold is first exceeded is identified as the pulsed edge moment; the second moment when the wheel travels to the preview obstacle is estimated according to the obstacle information; the time difference between the first moment and the second moment is calculated, and the time correction parameter is updated based on the time difference; the confidence level range corresponding to the obstacle identification confidence level is determined, and the obstacle identification confidence level is updated according to the update rule corresponding to the confidence level range.

[0087] In the embodiment of the present application, an approximate estimation of the actual time error of the pulsed edge event , is the moment when the pulsed road surface edge event occurs estimated according to the preview distance and the vehicle speed, is the actual pulsed road surface moment obtained by acceleration detection, and the value is used as new data to update the compensation value of the preview obstacle arrival time. Since the pre-mounted preview system is a time-varying system affected by factors such as light and vibration, a sliding window is used to limit the influence range of historical data, and the mean value within the time window is used as the reference value of the compensation value and the time window is setU After each new addition After: Mean update: (11) Let the time correction parameter (12) The suspension control for the rear wheels is judged according to the pulse road surface edge event. When the pulse road surface edge event occurs and the vehicle is in an alignable state with consistent bump and depression attributes, it is considered that the confidence of the preview obstacle information recognition is high, and the same control mode as the front wheels will be adopted for the rear wheel suspension.

[0088] One difference from the front wheels is that the confidence of the preview obstacle information will be updated.

[0089] (13) Another difference from the front wheels is that the processing method of the time difference is different. The front wheels process as shown in the calculation formula (12). For the rear wheels, since this alignment event has occurred in the front wheels, it can be directly obtained: (14) When the rear wheels pass through an obstacle i a hierarchical control strategy based on confidence is used to determine the working mode of the suspension. The hierarchical control strategy is consistent with the calculation logic of Module 3. Through the corrected input road surface elevation information and combined with its vehicle state sensors, the output force of the rear wheel suspension is calculated by the same control algorithm in Module 2 and Module 3.

[0090] If it cannot be aligned, there is no credible preview obstacle information for this pulse road surface event as a control reference, then the preview control module does not make an effective suspension force control for this pulse road surface event, and the suspension force is controlled by other algorithm modules.

[0091] In the embodiments of the present application, by detecting the root mean square values of acceleration in a relatively long window period and a relatively short window period, and judging whether the road surface input energy has a sudden change according to the ratio of the two root mean square values, the edge of the pulse road surface is detected, the time when the wheel travels onto the pulse road surface is judged, and the concave-convex property of the pulse road surface is judged according to the positive or negative of the mean value of the vertical acceleration of the unsprung mass in the short window period. According to the time and concave-convex property of the pulse road surface edge event, a comparison is made with the preview obstacle information. Considering the error of the preview obstacle, the pulse road surface edge within the time domain error range of a certain preview obstacle event and with the same concave-convex property is aligned with the preview obstacle in terms of space-time consistency. And the difference in the alignment event is recorded, the mean value of the difference in the alignment event is calculated, and the standard deviation of the difference is updated to provide a correction amount for the next obstacle event. For the preview obstacle event that can be aligned with the pulse road surface edge event, its confidence level is increased, and the preview obstacle information after alignment is used for the control of the rear wheel. If the pulse road surface edge event fails to be aligned with the preview recognized obstacle event, the inter-axle preview pulse road surface control method is adopted for the rear wheel.

[0092] In the embodiments of the present application, the edge of the pulse road surface is detected through an acceleration sensor, which can correct the preview recognition information in the current obstacle scenario and be used for the control of the rear wheel, and can also make a systematic correction to the preview recognition information in the subsequent obstacle scenario, so as to improve the accuracy of the input preview information and further improve the accuracy of the entire preview control system.

[0093] In another embodiment of the present application, as Figure 6 shown, a vehicle suspension control device is further provided, including: An acquisition module 11, configured to acquire the obstacle information and obstacle recognition confidence level of a preview obstacle obtained by performing preview recognition on the road surface on which the vehicle travels; A first determination module 12, configured to determine the optimal control force of an actuator of the suspension in the vehicle according to the obstacle information; A second determination module 13, configured to determine a target confidence interval to which the obstacle recognition confidence level belongs among a plurality of pre-set confidence intervals; A third determination module 14, configured to determine the actual control force of the suspension according to the target confidence interval and the optimal control force.

[0094] In another embodiment of the present application, an electronic device is further provided, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used for storing a computer program; The processor, when executing the program stored on the memory, implements the vehicle suspension control method described in any one of the foregoing method embodiments.

[0095] In the electronic device provided by an embodiment of the present invention, the processor determines the optimal control force of the actuator of the suspension by executing the program stored on the memory and using the obstacle information identified by previewing the road surface on which the vehicle travels, determines the target confidence interval in multiple preset confidence intervals by using the obstacle recognition confidence obtained by preview recognition, and then determines the actual control force of the suspension according to the target confidence interval and the optimal control force. By setting multiple candidate preset confidence intervals, hierarchical control of the actual output force of the suspension is realized according to different obstacle recognition confidences, so that the control of the actual output force of the suspension matches the obstacle recognition confidence, the control of the actual output force of the suspension more meets the actual driving conditions of the vehicle, improves the vehicle comfort, improves the smoothness when the vehicle passes through obstacles, and saves the energy consumption of suspension control.

[0096] The communication bus 1140 mentioned in the above electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus 1140 may be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 7 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0097] The communication interface 1120 is used for communication between the above electronic device and other devices.

[0098] The memory 1130 may include a Random Access Memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located far from the aforementioned processor.

[0099] The above-mentioned processor 1110 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0100] In another embodiment of the present application, a computer-readable storage medium is further provided. A program of a vehicle suspension control method is stored on the computer-readable storage medium. When the program of the vehicle suspension control method is executed by a processor, the steps of the vehicle suspension control method described in any of the foregoing method embodiments are implemented.

[0101] It should be noted that in this text, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device including the element.

[0102] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features claimed herein.

Claims

1. A vehicle suspension control method, characterized in that, Including: Obtaining obstacle information of a preview obstacle and an obstacle recognition confidence level obtained by performing preview recognition on a road surface where a vehicle travels; Determining an optimal control force of an actuator of a suspension in the vehicle according to the obstacle information; Determining a target confidence interval to which the obstacle recognition confidence level belongs among a plurality of preset confidence intervals; Determining an actual control force for the suspension according to the target confidence interval and the optimal control force.

2. The vehicle suspension control method according to claim 1, wherein The obstacle information includes: an obstacle longitudinal distance, an obstacle vertical height, an obstacle longitudinal width, and a time correction parameter. Determining the optimal control force of the actuator of the suspension in the vehicle according to the obstacle information includes: Determining a road surface elevation curve according to the obstacle longitudinal distance, the obstacle vertical height, and the obstacle longitudinal width; Converting the road surface elevation curve from a spatial domain into road surface excitation information in a time domain based on a current time correction parameter; Inputting the road surface excitation information into a 1 / 4 model of a vehicle suspension system to obtain suspension dynamic performance parameters; Determining an optimal control force of the actuator according to the suspension dynamic performance parameters and a preset optimal control algorithm.

3. The vehicle suspension control method according to claim 1, characterized in that, Determining the actual control force for the suspension according to the target confidence interval and the optimal control force includes: Determining a first suspension control mode for the suspension according to the target confidence interval and activating the first suspension control mode; Obtaining a current suspension speed of the suspension and determining a current maximum damping force and a current minimum damping force of a shock absorber of the suspension according to the current suspension speed; When the first suspension control mode is a non-silent control mode, determining the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force.

4. The vehicle suspension control method according to claim 3, wherein, If the target confidence interval is a first confidence interval and the first suspension control mode is a full active suspension control mode, determining the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force includes: When the optimal control force and the damping force of the suspension are in the same direction, if the optimal control force is less than or equal to the maximum damping force, controlling the shock absorber of the suspension to output a damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, controlling the shock absorber of the suspension to output the maximum damping force and controlling a front wheel actuator of the vehicle to output an actuating force according to a difference between the optimal control force and the maximum damping force; Or, when the optimal control force and the damping force of the suspension are in opposite directions, controlling the shock absorber of the suspension to output the minimum damping force and controlling a front wheel actuator of the vehicle to output an actuating force according to a sum value of the optimal control force and the maximum damping force.

5. The vehicle suspension control method according to claim 3, wherein If the target confidence interval is a second confidence interval, a maximum boundary of the second confidence interval is less than a minimum boundary of the first confidence interval, and the first suspension control mode is a semi-active suspension control mode; Determining the actual control force of the suspension based on the optimal control force, the maximum damping force, and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, control the shock absorber of the suspension to output the damping force according to the optimal control force; or, if the optimal control force is greater than the maximum damping force, control the shock absorber of the suspension to output the maximum damping force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, control the shock absorber of the suspension to output the minimum damping force.

6. The vehicle suspension control method according to claim 3, characterized in that, If the target confidence interval is the third confidence interval, and the maximum boundary of the third confidence interval is less than the minimum boundary of the second confidence interval, the first suspension control mode is a low attenuation semi-active suspension control mode; Determining the actual control force of the suspension based on the optimal control force, the maximum damping force and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the optimal control force and a preset confidence coefficient to obtain a first output force, and control the shock absorber of the suspension to output the damping force according to the first output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the maximum damping force and a preset confidence coefficient to obtain a second output force, and control the shock absorber of the suspension to output the damping force according to the second output force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, calculate the product of the minimum damping force and a preset confidence coefficient to obtain a third output force, and control the shock absorber of the suspension to output the damping force according to the third output force.

7. The vehicle suspension control method according to claim 3, wherein If the target confidence interval is the fourth confidence interval, and the maximum boundary of the fourth confidence interval is less than the minimum boundary of the third confidence interval, the first suspension control mode is a high attenuation semi-active suspension control mode; Determining the actual control force of the suspension of the vehicle based on the optimal control force, the maximum damping force and the minimum damping force includes: When the optimal control force is in the same direction as the damping force of the suspension, if the optimal control force is less than or equal to the maximum damping force, calculate the product of the square of the preset confidence coefficient and the optimal control force to obtain a fourth output force, and control the shock absorber of the suspension to output the damping force according to the fourth output force; or, if the optimal control force is greater than the maximum damping force, calculate the product of the square of the preset confidence coefficient and the maximum damping force to obtain a fifth output force, and control the shock absorber of the suspension to output the damping force according to the fifth output force; Or, when the optimal control force is in the opposite direction to the damping force of the suspension, calculate the product of the square of the preset confidence coefficient and the minimum damping force to obtain a sixth output force, and control the shock absorber of the suspension to output the damping force according to the sixth output force.

8. The vehicle suspension control method according to claim 1, wherein The method further includes: Obtain a plurality of first unsprung accelerations of the front wheel lower swing arm within a first time window; Determine whether the vehicle has traveled on a pulse road surface including pulse obstacles according to the plurality of first unsprung accelerations; If it is determined that the vehicle has traveled over a pulse road surface containing pulse obstacles, perform spatio-temporal consistency matching on the preview obstacle and the pulse obstacle according to the multiple first under-spring accelerations within the first time window and the obstacle information; If the spatio-temporal consistency matching between the preview obstacle and the pulse obstacle is successful, update the time correction parameter according to the multiple first under-spring accelerations within the first time window, and update the obstacle recognition confidence according to the preset update rule.

9. The vehicle suspension control method according to claim 8, wherein Determining whether the vehicle has traveled over a pulse road surface containing pulse obstacles according to the multiple first under-spring accelerations includes: Extract the first under-spring accelerations located within the second time window from the multiple first under-spring accelerations within the first time window, where the second time window is located within the first time window; Determine the first acceleration steady-state value within the long time window according to the first under-spring accelerations within the first time window; Determine the second acceleration steady-state value within the short time window according to the first under-spring accelerations within the second time window; Determine whether the ratio of the first acceleration steady-state value to the second acceleration steady-state value is greater than or equal to a preset steady-state threshold; If the ratio is greater than or equal to the preset threshold, determine that the front wheel of the vehicle has traveled over the pulse road surface.

10. The vehicle suspension control method according to claim 8, characterized in that, Performing spatio-temporal consistency matching on the preview obstacle and the pulse obstacle according to the multiple first under-spring accelerations within the first time window and the obstacle information includes: Determine the pulse edge moment when the wheel travels onto the pulse obstacle according to the multiple first under-spring accelerations within the first time window; If the pulse edge moment is within the preset time range determined according to the obstacle information and the dynamic error of the longitudinal distance of the obstacle conforms to a normal distribution, determine the first concave-convex attribute of the preview obstacle according to the obstacle information, and determine the second concave-convex attribute of the pulse obstacle according to the multiple first under-spring accelerations within the first time window; Determine whether the first concave-convex attribute is consistent with the second concave-convex attribute; If the first concave-convex attribute is consistent with the second concave-convex attribute, determine that the spatio-temporal consistency matching between the preview obstacle and the pulse obstacle passes.

11. The vehicle suspension control method according to claim 10, characterized in that Determining the first concave-convex attribute of the preview obstacle according to the obstacle information and the second concave-convex attribute of the pulse obstacle according to the multiple first under-spring accelerations within the first time window includes: Determine the first concave-convex attribute of the preview obstacle according to the obstacle type in the obstacle information; Calculate the acceleration mean value of the first under-spring accelerations within the second time window in the first time window; Determine the second concave-convex attribute of the pulse obstacle according to the positive or negative of the acceleration mean value.

12. The vehicle suspension control method according to claim 8, wherein, Updating the time correction parameter according to the multiple first under-spring accelerations within the first time window includes: Determine the first moment when the wheel travels onto the pulse obstacle according to the multiple first under-spring accelerations within the first time window; Estimate the second moment when the wheel travels onto the preview obstacle according to the obstacle information; Calculate the time difference between the first moment and the second moment, and update the time correction parameter based on the time difference; Determine the confidence range corresponding to the obstacle recognition confidence, and update the obstacle recognition confidence according to the update rule corresponding to the confidence range.

13. A vehicle suspension control device, characterized in that, Including: An acquisition module, configured to acquire obstacle information and obstacle recognition confidence of a preview obstacle obtained by performing preview recognition on a road surface where a vehicle travels; A first determination module, configured to determine an optimal control force of an actuator of a suspension in the vehicle according to the obstacle information; A second determination module, configured to determine a target confidence interval to which the obstacle recognition confidence belongs among a plurality of preset confidence intervals; A third determination module, configured to determine an actual control force of the suspension according to the target confidence interval and the optimal control force.

14. An electronic device, characterized in that, Including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete communication with each other through the communication bus; The memory is used for storing a computer program; The processor is configured to implement the vehicle suspension control method according to any one of claims 1-12 when executing the program stored on the memory.

15. A computer-readable storage medium, characterized in that, A program of the vehicle suspension control method is stored on the computer-readable storage medium, and when the program of the vehicle suspension control method is executed by the processor, the steps of the vehicle suspension control method according to any one of claims 1-12 are implemented.

Images