Vehicle control method and apparatus, electronic device, and medium

Abstract

The present application relates to the technical field of vehicles, and provides a vehicle control method and apparatus, an electronic device, and a medium. The method comprises: on the basis of the transverse motion state quantity and the longitudinal motion state quantity of a first vehicle, solving for a tracking control problem for tracking a planning path and a planning speed, so as to obtain a first wheel steering angle control quantity and an expected acceleration of the first vehicle, wherein the first wheel is a front wheel or a rear wheel, the planning path comprises a reference value of the transverse motion state quantity of the vehicle at a series of time points in the future, and the planning speed comprises a reference value of the longitudinal motion state quantity of the vehicle at a series of time points in the future; obtaining the acceleration control quantity of the first vehicle on the basis of the expected acceleration of the first vehicle and the actual acceleration thereof; and controlling the first vehicle on the basis of the first wheel steering angle control quantity and the acceleration control quantity of the first vehicle. In the present application, vehicle control is carried out in a transverse and longitudinal combined tracking control mode, so that the speed tracking accuracy can be improved as much as possible while the path tracking accuracy is ensured.

Description

Vehicle control methods, devices, electronic equipment and media

[0001] This application claims priority to Chinese Patent Application No. 202410488311.0, filed on April 22, 2024, entitled "Vehicle Control Method, Apparatus, Electronic Device and Medium", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of vehicle technology, and in particular to a vehicle control method, device, electronic device, and medium. Background Technology

[0003] The fully automatic vehicle control system can achieve lateral tracking control based on the planned path by controlling the vehicle's steering mechanism, and longitudinal tracking control based on the planned speed by controlling the vehicle's drive and braking mechanism. This can effectively solve the problem of driving difficulties for inexperienced drivers in complex spaces.

[0004] Since the vehicle is in a low-speed, large-angle motion state during driving, the longitudinal and lateral forces of the tires are both in the linear working range. Therefore, a lateral and longitudinal decoupled tracking control method can be used to control the steering mechanism and the drive and braking mechanism to achieve lateral and longitudinal tracking control of the vehicle.

[0005] However, when the curvature of the planned path changes at a large rate, the inherent speed constraint of the vehicle steering mechanism requires the vehicle to actively reduce its speed in order to achieve accurate tracking control of the planned path. This poses a significant challenge to the existing lateral and longitudinal decoupled tracking control method. Summary of the Invention

[0006] This application provides a vehicle control method, device, electronic device, and medium that uses a combined lateral and longitudinal tracking control method to control the vehicle, which can improve speed tracking accuracy as much as possible while ensuring path tracking accuracy.

[0007] In a first aspect, embodiments of this application provide a vehicle control method, comprising: solving a tracking control problem for tracking a planned path and a planned speed based on the lateral motion state variables and longitudinal motion state variables of a first vehicle, to obtain a first wheel steering angle control variable and a desired acceleration of the first vehicle; wherein the first wheel is a front wheel or a rear wheel, the planned path includes reference values ​​of the vehicle's lateral motion state variables at a series of future time points, and the planned speed includes reference values ​​of the vehicle's longitudinal motion state variables at a series of future time points; obtaining an acceleration control variable of the first vehicle based on the desired acceleration and the actual acceleration of the first vehicle; and controlling the first vehicle based on the first wheel steering angle control variable and the acceleration control variable.

[0008] This application solves the joint lateral and longitudinal tracking control problem for tracking the planned path and speed together based on the current lateral and longitudinal motion state of the vehicle. This allows the lateral control quantity and the desired longitudinal control quantity to meet the requirements of joint lateral and longitudinal tracking, thereby realizing the lateral and longitudinal control of the vehicle. This can improve the speed tracking accuracy as much as possible while ensuring the path tracking accuracy.

[0009] Optionally, the tracking control problem includes vehicle motion equations; the state variables of the vehicle motion equations include at least one of the following: lateral position, heading angle, yaw angle, lateral velocity, heading angular velocity, yaw angular velocity, and first wheel steering angle; and at least one of the following: longitudinal displacement, longitudinal velocity, and acceleration; the control variables of the vehicle motion equations include: first wheel steering angle control variable and desired acceleration.

[0010] By limiting the tracking control problem to include the combined lateral and longitudinal motion equations of the vehicle, the solution results can satisfy both the lateral and longitudinal motion equations of the vehicle, thereby enabling the vehicle to accurately track the planned path laterally and the planned speed longitudinally based on the solution results.

[0011] Optionally, the tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimize the first function; the first function includes path tracking error and speed tracking error, wherein the path tracking error is obtained from the error between the vehicle's lateral motion state quantity and the reference value in the planned path, and the speed tracking error is obtained from the error between the vehicle's longitudinal motion state quantity and the reference value in the planned speed.

[0012] By defining the tracking control problem as including a first function involving lateral and longitudinal tracking errors, and by finding the solution to the tracking control problem as the data that minimizes the first function, the tracking error of the vehicle when controlling the vehicle based on the solution meets the requirements, thereby achieving accurate tracking of lateral and longitudinal planning information.

[0013] Optionally, the tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimize the first function; the first function includes process constraint control quantity and a first weight of the process constraint control quantity, terminal constraint control quantity and a second weight of the terminal constraint control quantity; wherein, the first weight and the second weight are obtained according to the scenario in which the first vehicle is located; the terminal constraint control quantity includes the path tracking error and speed tracking error at the future target time point; the process constraint control quantity includes the path tracking error and speed tracking error at each time point within the future time period, and the target time point is the next time point of the future time period.

[0014] By limiting the first function in the tracking control problem to include process constraint control quantities and terminal constraint control quantities, the solution to the tracking control problem not only conforms to the lateral and longitudinal planning information in the prediction process, but also conforms to the lateral and longitudinal planning information at the prediction endpoint, enabling the vehicle to accurately track the lateral and longitudinal planning information based on the solution results.

[0015] Furthermore, by combining the weights of the lateral and longitudinal tracking targets to adjust the process constraint control quantity and its proportion in the first function, the first function can be adapted to the current vehicle driving scenario, thereby supporting accurate lateral and longitudinal tracking driving in various vehicle driving scenarios.

[0016] Optionally, the tracking control problem is a linear constraint optimization problem including the first function, and the constraints of the linear constraint optimization problem include vehicle lateral control constraints and vehicle longitudinal control constraints; solving the tracking control problem for tracking the planned path and planned speed includes: solving the linear constraint optimization problem using the primal dual interior point method.

[0017] By transforming the lateral and longitudinal joint tracking control problem into a linear constraint optimization problem with lateral and longitudinal control constraints, and using the primal dual interior point method to solve the linear constraint optimization problem, the solution results of the lateral and longitudinal joint control that meet the lateral and longitudinal control constraints can be obtained. Based on the solution results, vehicle lateral and longitudinal tracking control that meets the lateral and longitudinal control constraints can be realized.

[0018] Optionally, the vehicle control method further includes: using a rolling time-domain estimation module with a time-domain length of 1 and a rolling time-domain estimation module with a time-domain length of N to estimate the actual acceleration of the first vehicle, where N > 1; the rolling time-domain estimation module with a time-domain length of 1 is used to obtain a first parameter based on the measured value of the longitudinal velocity of the first vehicle in a first time period when the number of time points of the vehicle's travel time is greater than N; the rolling time-domain estimation module with a time-domain length of N is used to estimate the actual acceleration of the first vehicle based on the first parameter and the measured value of the longitudinal velocity of the first vehicle in a second time period when the number of time points of the vehicle's travel time is greater than N; wherein, the second time period is connected to the first time period, and the starting point of the first time period is the initial travel time of the first vehicle.

[0019] By employing a rolling time domain that suppresses information transmission delays to estimate vehicle acceleration, the actual vehicle acceleration that satisfies physical constraints of state variables (such as vehicle speed constraints) can be estimated in real time and accurately. Using the acceleration control variable generated based on the accurately estimated actual vehicle acceleration for longitudinal vehicle control helps improve the accuracy of longitudinal tracking control.

[0020] The actual acceleration of the vehicle is estimated by running a rolling time-domain estimation module with a time-domain length of 1 and a rolling time-domain estimation module with a time-domain length of N in parallel. This ensures the timeliness of the estimation while taking into account the impact of all historical vehicle speeds on the real-time estimation of vehicle acceleration, thereby improving the accuracy of the estimation of the actual vehicle acceleration.

[0021] Optionally, obtaining the acceleration control quantity of the first vehicle based on the desired acceleration and the actual acceleration of the first vehicle includes: inputting the desired acceleration and the actual acceleration of the first vehicle into a proportional-integral-derivative (PID) controller to obtain the acceleration control quantity of the first vehicle output by the PID controller.

[0022] By designing an inner-loop control strategy for vehicle acceleration, and generating acceleration control quantities based on accurately estimated vehicle body acceleration and the desired acceleration obtained by solving the joint longitudinal and lateral tracking control problem, it is helpful to achieve precise longitudinal tracking control of the vehicle.

[0023] Secondly, embodiments of this application provide a vehicle control device, comprising: a first processing module, configured to solve a tracking control problem for tracking a planned path and a planned speed based on the lateral motion state quantity and longitudinal motion state quantity of a first vehicle, and obtain a first wheel steering angle control quantity and a desired acceleration of the first vehicle; wherein the first wheel is a front wheel or a rear wheel, the planned path includes reference values ​​of the lateral motion state quantity of the vehicle at a series of future time points, and the planned speed includes reference values ​​of the longitudinal motion state quantity of the vehicle at a series of future time points; a second processing module, configured to obtain an acceleration control quantity of the first vehicle based on the desired acceleration of the first vehicle and the actual acceleration of the first vehicle; and a control module, configured to control the first vehicle based on the first wheel steering angle control quantity and the acceleration control quantity.

[0024] Thirdly, embodiments of this application provide a chip, including: a processor, which is configured to execute computer program instructions stored in a memory, wherein when the computer program instructions are executed by the processor, the chip is triggered to execute the method as described in any of the first aspects.

[0025] Fourthly, embodiments of this application provide an electronic device including at least one processor and a memory coupled together. The memory is used to store computer program instructions, and the processor is used to execute the computer program instructions. When the computer program instructions are executed by the processor, the electronic device is triggered to perform a method as described in any of the first aspects.

[0026] Fifthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the method as described in any of the first aspects.

[0027] In a sixth aspect, embodiments of this application provide a computer program product, which includes a computer program that, when run on a computer, causes the computer to perform the method as described in any of the first aspects.

[0028] The technical effects of the aforementioned aspects can be referenced from each other, and will not be elaborated further here. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below.

[0030] Figure 1 is a schematic diagram of a combined longitudinal and transverse tracking control system provided in an embodiment of this application;

[0031] Figure 2 is a schematic flowchart of a vehicle control method provided in an embodiment of this application;

[0032] Figure 3 is a schematic diagram of a vehicle lateral tracking control based on a planned path provided in an embodiment of this application;

[0033] Figure 4 is a schematic diagram of a vehicle acceleration estimation based on rolling time domain provided in an embodiment of this application;

[0034] Figure 5 is a block diagram of a vehicle control device provided in an embodiment of this application;

[0035] Figure 6 is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0036] To better understand the technical solution of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0037] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0038] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0039] It should be understood that the term "at least one" as used in this document refers to one or more, and "more than one" refers to two or more. The term "and / or" as used in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. A and B can be singular or plural. Additionally, the character " / " in this document generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0040] It should be understood that although the terms "first," "second," etc., may be used to describe the set thresholds in the embodiments of this application, these set thresholds should not be limited to these terms. These terms are only used to distinguish the set thresholds from each other. For example, without departing from the scope of the embodiments of this application, the first set threshold may also be referred to as the second set threshold, and similarly, the second set threshold may also be referred to as the first set threshold.

[0041] Fully automated vehicle control systems can achieve lateral tracking control based on a planned path by controlling the vehicle's steering mechanism, and longitudinal tracking control based on a planned speed by controlling the vehicle's drive and braking mechanisms. This effectively solves the problem of inexperienced drivers struggling to navigate complex spaces. For example, fully automated parking systems can achieve parking path tracking control by controlling the vehicle's steering mechanism and parking speed tracking control by controlling the vehicle's drive and braking mechanisms, thus solving the problem of inexperienced drivers struggling to park in narrow spaces.

[0042] Since vehicles are in a low-speed, large-angle motion state during driving (such as parking), the longitudinal and lateral forces of the tires are both in the linear working range. Therefore, a lateral and longitudinal decoupled tracking control method (i.e., lateral control and longitudinal control are independent of each other and do not affect each other) can be used to control the steering mechanism and drive braking mechanism to achieve lateral and longitudinal tracking control of the vehicle.

[0043] However, when the curvature of the planned path changes at a large rate, the inherent speed constraint of the vehicle steering mechanism requires the vehicle to actively reduce its speed in order to achieve accurate tracking control of the planned path. This poses a significant challenge to the existing lateral and longitudinal decoupled tracking control method.

[0044] To achieve precise tracking control in both the lateral and longitudinal directions simultaneously, this application provides a vehicle control strategy that employs a combined lateral and longitudinal tracking control method, which can maximize speed tracking accuracy while ensuring path tracking accuracy.

[0045] Referring to Figure 1, a lateral and longitudinal joint tracking control system is shown. The system includes a lateral and longitudinal joint tracking control module 101, a proportional-integral-derivative (PID) controller 102, and a rolling time-domain-based acceleration estimation module 103. The system can include two parts: an outer-loop control and an inner-loop control.

[0046] In the outer loop control, the horizontal and vertical joint tracking control module 101 follows the planned path (including, for example, y). ref , Reference values ​​for vehicle lateral motion states and planned speeds (including, for example, a) ref Reference values ​​for the vehicle's longitudinal motion state (e.g., a0, y0), and the vehicle's current lateral and longitudinal motion state (e.g., a0, y0, ...). Solve the joint lateral and longitudinal tracking control problem to obtain the joint lateral and longitudinal control variables, which include the lateral motion control variable δ. d and longitudinal motion control quantity a d Where y represents the lateral position and δ represents the wheel steering angle. α represents the yaw angle, α represents the acceleration, and v represents the acceleration. x This indicates longitudinal velocity.

[0047] Referring to Figure 1, the lateral motion control quantity δ can be used as a reference. d The steering mechanism of the vehicle body 100 is controlled to achieve lateral control of the vehicle. The longitudinal motion control quantity a d It can be used as the desired acceleration to obtain the acceleration control quantity a.

[0048] In the inner-loop control, the acceleration estimation module 103 based on the rolling time domain estimates all historical longitudinal velocities v measured by the vehicle speed sensor. x Estimate the vehicle's current actual acceleration The proportional-integral-derivative controller 102 inputs the estimated actual acceleration. and longitudinal motion control quantity a d The output acceleration control quantity 'a' is used. The current acceleration used in the outer loop solution of the tracking control problem can be the actual acceleration estimated by the acceleration estimation module 103.

[0049] Referring to Figure 1, the drive braking mechanism of the vehicle body 100 can be controlled according to the acceleration control amount a to achieve lateral control of the vehicle.

[0050] In the embodiment shown in Figure 1, the outer loop solves the joint lateral and longitudinal tracking control problem to obtain the joint lateral and longitudinal control quantities. Based on this, vehicle control can ensure that speed tracking accuracy is improved as much as possible while maintaining path tracking accuracy.

[0051] The distributed electronic and electrical architecture of mainstream vehicles causes transmission delays in the wheel speed and vehicle acceleration information measured by sensors. Furthermore, when estimating vehicle acceleration based on Kalman filtering, the influence of physical constraints on state variables cannot be considered, which affects the smoothness and accuracy of speed tracking control based on vehicle acceleration closed loop.

[0052] In the embodiment shown in Figure 1, the inner loop uses a rolling time domain that can suppress information transmission delay, and estimates the vehicle acceleration that meets the physical constraints of the state variables in real time. Combined with a PID controller, it can achieve precise tracking control of the planned speed, which can improve the smoothness and accuracy of speed tracking control.

[0053] As shown in Figure 2, this application embodiment provides a vehicle control method, which may include the following steps 201 to 203:

[0054] Step 201: Based on the lateral and longitudinal motion state variables of the first vehicle, solve the tracking control problem for tracking the planned path and planned speed to obtain the first wheel steering angle control variable and the desired acceleration of the first vehicle. Here, the first wheel is either the front wheel or the rear wheel, the planned path includes reference values ​​of the vehicle's lateral motion state variables at a series of future time points, and the planned speed includes reference values ​​of the vehicle's longitudinal motion state variables at a series of future time points.

[0055] In one embodiment, the first wheel is the vehicle's drive wheel.

[0056] For example, the lateral motion state variables of the first vehicle may include lateral motion parameters such as the first vehicle's current lateral (or sideways) position, heading angle or yaw angle, lateral velocity, heading angular velocity or yaw angular velocity, and wheel steering angle. Correspondingly, the planned path may contain reference values ​​for these lateral motion parameters; for example, the planned path may include the lateral position at a series of future time points. The vehicle's yaw angle can be the difference between the heading angle and the sideslip angle.

[0057] For example, the longitudinal motion state variables of the first vehicle may include longitudinal motion parameters such as the current longitudinal displacement (distance between longitudinal positions at two adjacent time points), longitudinal velocity, and acceleration. Correspondingly, the planned speed may include reference values ​​for these longitudinal motion parameters; for example, the planned speed may include longitudinal velocities at a series of future time points.

[0058] Feasibly, in solving the tracking control problem, the control quantities to be executed by the vehicle at the current time step (i.e., lateral and longitudinal control quantities) can be predicted based on the error between the vehicle's current state quantities (i.e., lateral and longitudinal state quantities) and the corresponding planned state quantities. Then, based on the current state quantities and the control quantities to be executed, the vehicle's state quantity at the next time step can be predicted. Finally, based on the error between the vehicle's state quantity at the next time step and the corresponding planned state quantities, the control quantity to be executed at the next time step can be predicted. This process can be repeated to predict the vehicle's state quantities and control quantities at a series of future time points. If this series of state quantities meets the error requirement compared to the planned state quantities, the currently executed control quantities can be used to control the vehicle, achieving precise tracking.

[0059] As can be seen, the embodiment shown in Figure 2 solves the joint tracking control problem of the lateral and longitudinal directions for tracking the planned path and speed together by based on the current lateral and longitudinal motion state of the vehicle. This can obtain the lateral control quantity and the longitudinal desired control quantity that meet the requirements of joint tracking, so as to realize the lateral and longitudinal control of the vehicle. In this way, the speed tracking accuracy can be improved as much as possible while ensuring the path tracking accuracy.

[0060] In one embodiment of this application, the tracking control problem includes a vehicle motion equation; the state variables of the vehicle motion equation include at least one of the following: lateral position, heading angle, yaw angle, lateral velocity, heading angular velocity, yaw angular velocity, and first wheel steering angle; and at least one of the following: longitudinal displacement, longitudinal velocity, and acceleration; the control variables of the vehicle motion equation include: first wheel steering angle control variable and desired acceleration.

[0061] Feasibly, in order to achieve lateral and longitudinal joint control, the vehicle motion equations in the lateral and longitudinal joint tracking control problem can be the vehicle lateral and longitudinal joint motion equations obtained from the vehicle longitudinal motion equations and the vehicle lateral motion equations, in order to solve for the vehicle lateral and longitudinal control quantities that conform to the vehicle lateral motion equations and the vehicle longitudinal motion equations, and to ensure the effectiveness of the solution results.

[0062] Practically, the vehicle lateral motion equation can be used to describe the relationship between the vehicle's lateral motion state variables and the vehicle's lateral motion control variables. The vehicle's lateral motion state variables can include lateral position, yaw angle, lateral velocity, yaw rate, front wheel steering angle, etc., while the vehicle's lateral motion control variables can be the front wheel steering angle control variables.

[0063] In one feasible implementation, the relationship between state variables and control variables can be constructed in the vehicle's lateral motion equation based on the inertial element time constant of the vehicle's steering system response characteristics and the vehicle's structural parameters (such as the lateral stiffness of the front and rear wheels, the distance from the vehicle's center of gravity to the front and rear axles, etc.).

[0064] Feasibly, the vehicle longitudinal motion equation can be used to describe the relationship between the vehicle's longitudinal motion state variables and the vehicle's longitudinal motion control variables. The vehicle's longitudinal motion state variables can include longitudinal displacement, longitudinal velocity, acceleration, etc., while the vehicle's longitudinal motion control variable can be the desired acceleration.

[0065] In one feasible implementation, the relationship between state variables and control variables can be constructed in the vehicle's longitudinal motion equation based on the time constant of the inertial element of the vehicle's drive and braking system response characteristics.

[0066] By limiting the tracking control problem to include the combined lateral and longitudinal motion equations of the vehicle, the solution results can satisfy both the lateral and longitudinal motion equations of the vehicle, thereby enabling the vehicle to accurately track the planned path laterally and the planned speed longitudinally based on the solution results.

[0067] In one embodiment of this application, the tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimize the first function; the first function includes path tracking error and speed tracking error, wherein the path tracking error is obtained based on the error between the vehicle's lateral motion state quantity and the reference value in the planned path, and the speed tracking error is obtained based on the error between the vehicle's longitudinal motion state quantity and the reference value in the planned speed.

[0068] In one feasible implementation, the velocity tracking error can include longitudinal displacement error, longitudinal velocity error, acceleration error, etc.

[0069] In one feasible implementation, path tracking errors can include lateral position errors, heading angle errors, etc.

[0070] By solving for the control quantity that minimizes the first function, the error between the vehicle's lateral and longitudinal driving states and the planned corresponding state quantities can be minimized when controlling the vehicle based on the solved control quantity, thereby improving the accuracy of tracking.

[0071] Thus, by limiting the tracking control problem to include a first function involving lateral and longitudinal tracking errors, and by ensuring that the solution to the tracking control problem is the data that minimizes the first function, the tracking error of the vehicle when controlling the vehicle based on the solution meets the requirements, thereby achieving accurate tracking of lateral and longitudinal planning information.

[0072] In one embodiment of this application, the tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimizes the first function; the first function includes process constraint control quantity and terminal constraint control quantity; the terminal constraint control quantity includes the path tracking error and speed tracking error at the future target time point; the process constraint control quantity includes the path tracking error and speed tracking error at each time point within the future time period, and the target time point is the next time point of the future time period.

[0073] In solving the tracking control problem, the state variables at a series of future time points can be predicted. Based on the state variables at the final predicted time point, the tracking error between this state variable and the corresponding reference value can be used as the terminal constraint control variable. Conversely, based on the state variables at intermediate predicted time points, the tracking error between this state variable and the corresponding reference value can be used as the process constraint control variable. These two control variables can reflect the predictive tracking effect from two different perspectives: process and termination.

[0074] Thus, by limiting the first function in the tracking control problem to include process constraint control quantities and terminal constraint control quantities, the solution to the tracking control problem not only conforms to the lateral and longitudinal planning information in the prediction process, but also conforms to the lateral and longitudinal planning information at the prediction endpoint, supporting the vehicle to accurately track the lateral and longitudinal planning information based on the solution results.

[0075] In one embodiment of this application, the first function includes a process constraint control quantity and a first weight of the process constraint control quantity, a terminal constraint control quantity and a second weight of the terminal constraint control quantity; wherein the first weight and the second weight are obtained according to the scenario in which the first vehicle is located.

[0076] In one embodiment, a mapping relationship between the vehicle driving scenario and the joint weights of the horizontal and vertical axes can be preset, and this mapping relationship can be obtained based on experience.

[0077] For example, vehicle driving scenarios can include automatic parking scenarios, highway driving scenarios, highway navigation driving, urban navigation driving, and autonomous valet parking, among other autonomous driving scenarios.

[0078] By using lateral and longitudinal joint weights adapted to the vehicle driving scenario, a lateral and longitudinal joint tracking control problem can be constructed, and vehicle control can be performed based on the solution results, thereby achieving a tracking control effect that conforms to the vehicle driving scenario.

[0079] For example, in high-speed driving scenarios, the longitudinal and lateral forces of the tires are in a coupled operating region. If a lateral-longitudinal decoupled tracking control method is used, it is difficult to ensure that the longitudinal and lateral forces satisfy the tire friction ellipse constraint, which can easily lead to skidding or insufficient braking. However, the embodiments of this application use a lateral-longitudinal joint tracking control method and adopt lateral-longitudinal joint weights adapted to high-speed driving scenarios to construct a lateral-longitudinal joint tracking control problem. Based on this, the tire friction ellipse constraint problem can be better solved when controlling the vehicle.

[0080] By combining the weights of the lateral and longitudinal tracking targets to adjust the process constraint control quantity and its proportion in the first function, the first function can be adapted to the current vehicle driving scenario, thereby supporting accurate lateral and longitudinal tracking driving in various vehicle driving scenarios.

[0081] In one embodiment of this application, the tracking control problem is a linear constraint optimization problem including a first function, and the constraints of the linear constraint optimization problem include vehicle lateral control constraints and vehicle longitudinal control constraints; solving the tracking control problem for tracking the planned path and planned speed includes: solving the linear constraint optimization problem using the primal dual interior point method.

[0082] In one embodiment, the vehicle lateral control constraints can be constructed based on the maximum and minimum values ​​of the wheel steering angle, in order to solve for the wheel steering angle control quantity that satisfies the wheel steering angle constraints, thereby ensuring the effectiveness of the vehicle lateral control.

[0083] In one embodiment, longitudinal control constraints for the vehicle can be constructed based on the maximum and minimum values ​​of vehicle acceleration, in order to solve for the vehicle acceleration constraints and the desired acceleration, thereby ensuring the effectiveness of the vehicle longitudinal control.

[0084] After transforming the tracking control problem into a linear constraint optimization problem based on the horizontal and vertical constraints, the primal dual interior point method can be used to solve the linear constraint optimization problem to obtain a solution that satisfies the joint horizontal and vertical control requirements.

[0085] Feasibly, in the process of solving linear constrained optimization problems using the primal dual interior point method, a Lagrange function and slack variables can be introduced, so that the solution can be obtained by iteratively applying parameters such as Lagrange multipliers and slack variables.

[0086] It is evident that by transforming the joint lateral and longitudinal tracking control problem into a linear constraint optimization problem with lateral and longitudinal control constraints, and by using the primal dual interior point method to solve the linear constraint optimization problem, a solution for the joint lateral and longitudinal control that conforms to the lateral and longitudinal control constraints can be obtained. Furthermore, based on the solution, vehicle lateral and longitudinal tracking control that conforms to the lateral and longitudinal control constraints can be achieved.

[0087] Step 202: Obtain the acceleration control amount of the first vehicle based on the desired acceleration of the first vehicle and the actual acceleration of the first vehicle.

[0088] Feasibly, on-board sensors can be used to measure the longitudinal velocity of the first vehicle in real time (e.g., by measuring wheel speed), and then the actual acceleration of the first vehicle can be estimated in real time based on the measured longitudinal velocity.

[0089] In one embodiment of this application, step 202 may include: inputting the desired acceleration of the first vehicle and the actual acceleration of the first vehicle into a proportional-integral-derivative (PID) controller to obtain the acceleration control quantity of the first vehicle output by the PID controller.

[0090] After obtaining the desired acceleration, the desired acceleration and the vehicle's current actual acceleration can be input into a PID controller. The acceleration output by the PID controller is then used as the acceleration controller to control the vehicle's drive and braking mechanisms to achieve longitudinal control. Based on the control effect of the PID controller, the error between theoretical acceleration control and actual control can be reduced, improving the stability of the vehicle's longitudinal control.

[0091] By designing an inner-loop control strategy for vehicle acceleration, and generating acceleration control quantities based on accurately estimated vehicle body acceleration and the desired acceleration obtained by solving the joint longitudinal and lateral tracking control problem, it is helpful to achieve precise longitudinal tracking control of the vehicle.

[0092] Step 203: Control the first vehicle according to the first wheel steering angle control amount and acceleration control amount of the first vehicle.

[0093] Taking the front wheels as an example, solving the combined lateral and longitudinal tracking control problem yields the lateral motion control quantity of the front wheel steering angle. This front wheel steering angle control quantity can then be used to control the vehicle's steering mechanism for lateral tracking control. Compared to the rear wheel steering angle control module, the front wheel steering angle control module is independent of the vehicle chassis interface. Therefore, by designing a high-performance control law for combined lateral and longitudinal tracking control, the control performance requirements can be met, enabling precise path tracking control.

[0094] If the vehicle is a four-wheel steering car, solving the tracking control problem can yield the front wheel steering angle control value. Then, based on the front wheel steering angle control value and the correlation between the front and rear wheel steering angles, the rear wheel steering angle control value can be obtained. The lateral tracking control of the four-wheel steering car can then be performed based on the front wheel steering angle control value and the rear wheel steering angle control value.

[0095] Solving the combined lateral and longitudinal tracking control problem also yields the longitudinal motion control quantity. Since the acceleration compensation module does not rely on the vehicle chassis interface, and the speed tracking control module is independent of the vehicle chassis interface, designing a high-performance control law for combined lateral and longitudinal tracking control can meet the control performance requirements and achieve precise speed tracking control.

[0096] To balance the control performance and robustness of vehicle control, instead of directly using the solved longitudinal motion control quantity to control vehicle acceleration, the longitudinal motion control quantity is used as the desired acceleration, and the current estimated actual acceleration of the vehicle is used for acceleration compensation before longitudinal vehicle control is performed.

[0097] The distributed electronic and electrical architecture of mainstream vehicles introduces transmission delays in wheel speed and vehicle acceleration information measured by sensors. Furthermore, Kalman filtering-based vehicle acceleration estimation cannot account for the influence of physical constraints on state variables, thus affecting the smoothness and accuracy of speed tracking control based on vehicle acceleration closed-loop. Therefore, a rolling time domain, which suppresses information transmission delays, can be used to estimate the actual vehicle acceleration in real time, thereby improving the accuracy of vehicle acceleration estimation.

[0098] In one embodiment of this application, the vehicle control method further includes: estimating the actual acceleration of the first vehicle using a rolling time-domain estimation module with a time-domain length of 1 and a rolling time-domain estimation module with a time-domain length of N, where N > 1.

[0099] The rolling time-domain estimation module with a time-domain length of 1 is used to obtain a first parameter based on the measured longitudinal velocity of the first vehicle in the first time period when the number of time points of the vehicle's travel time is greater than N. The rolling time-domain estimation module with a time-domain length of N is used to estimate the actual acceleration of the first vehicle based on the first parameter and the measured longitudinal velocity of the first vehicle in the second time period when the number of time points of the vehicle's travel time is greater than N. The second time period is connected to the first time period, and the starting point of the first time period is the initial travel time of the first vehicle.

[0100] Feasibly, the end point of the second time period is the previous time, that is, the actual acceleration of the vehicle at the current time can be estimated based on the longitudinal velocities of all vehicles before the current time.

[0101] In one feasible implementation, the historical measurement data used by the rolling time-domain estimation module may include not only the vehicle's longitudinal velocity but also the vehicle's acceleration, which can be obtained by differentiating the vehicle's longitudinal velocity.

[0102] For example, N can be 10, then a rolling time-domain estimation module with a time-domain length of 10 can use measurements at 10 time points.

[0103] If the number of time points in the vehicle's travel time is no greater than N, then the rolling time domain estimation module with a time domain length of N can use all historical measurements of the vehicle. Therefore, it can estimate the vehicle's current actual acceleration based solely on this rolling time domain, according to all historical longitudinal velocities of the vehicle.

[0104] If the number of time points in the vehicle's driving time is greater than N, the rolling time-domain estimation module with a time-domain length of N cannot use all the vehicle's historical measurement values. Therefore, the rolling time-domain estimation module with a time-domain length of N and the rolling time-domain estimation module with a time-domain length of 1 can be combined to estimate the vehicle's current actual acceleration based on all the vehicle's historical longitudinal velocities, etc.

[0105] Taking N=10 as an example, if the current time is t, the rolling time domain estimation module with a time domain length of 1 can use the vehicle's longitudinal velocity in the time period from 0 to t-11, while the rolling time domain with a time domain length of 10 can use the vehicle's longitudinal velocity in the time period from t-10 to t-1. Thus, based on all the vehicle's historical longitudinal velocities, the actual acceleration of the vehicle at time t can be estimated.

[0106] If the current time is t+1, the rolling time-domain estimation module with a time-domain length of 1 can use the vehicle's longitudinal velocity within the time period from 0 to t-10, while the rolling time-domain module with a time-domain length of 10 can use the vehicle's longitudinal velocity within the time period from t-9 to t. This allows for the estimation of the vehicle's actual acceleration at time t based on all its historical longitudinal velocities. This process continues to achieve a real-time and accurate estimation of the vehicle's actual acceleration. It is evident that the value ranges of the two rolling time-domain modules change in real-time over time.

[0107] Feasibly, taking the current time t as an example, the rolling time-domain estimation module with a time-domain length of 1 can reassign the vehicle's longitudinal velocity at time point tN used by the rolling time-domain estimation module with a time-domain length of N based on the vehicle's longitudinal velocity in the time period from 0 to tN-1. This ensures that the actual vehicle acceleration estimated by the rolling time-domain estimation module with a time-domain length of N is affected not only by the vehicle's longitudinal velocity in the time period from t-1 to tN, but also by the vehicle's longitudinal velocity in the time period from 0 to tN-1, thereby achieving the effect of accurately estimating the vehicle's current acceleration based on the historical vehicle speeds of all vehicles.

[0108] Compared to implementations that estimate vehicle acceleration solely based on current speed, this approach avoids using a large amount of historical vehicle speed data, thus preserving estimation accuracy. Furthermore, by combining two rolling time domains and real-time rolling, a vast amount of historical vehicle speed data can be utilized to estimate vehicle acceleration, significantly improving the accuracy of acceleration estimation.

[0109] Compared to methods that directly use a large amount of historical vehicle speed data to estimate the vehicle's current acceleration, this approach involves excessive data processing, thus affecting the timeliness of the acceleration estimation. The inherent characteristic of a rolling time-domain estimation module with a time-domain length of 1 is its low computational power. Furthermore, a rolling time-domain estimation module with a time-domain length of N has both low data volume and low computational power. Therefore, by combining the two rolling time-domain methods and implementing real-time rolling, both the accuracy and timeliness of the estimation can be guaranteed.

[0110] It is evident that real-time scrolling based on a scrolling time domain of length N ensures the timeliness of acceleration estimation, while real-time scrolling based on a scrolling time domain of length 1 ensures the accuracy of acceleration estimation. Combining these two scrolling time domains allows for a balance between the timeliness and accuracy of acceleration estimation.

[0111] By employing a rolling time domain that suppresses information transmission delays to estimate vehicle acceleration, the actual vehicle acceleration satisfying the physical constraints of the state variables can be estimated in real time and accurately. Using the acceleration control variable generated based on the accurately estimated actual vehicle acceleration for longitudinal vehicle control helps improve the accuracy of longitudinal tracking control.

[0112] The actual acceleration of the vehicle is estimated by running a rolling time-domain estimation module with a time-domain length of 1 and a rolling time-domain estimation module with a time-domain length of N in parallel. This ensures the timeliness of the estimation while taking into account the impact of all historical vehicle speeds on the real-time estimation of vehicle acceleration, thereby improving the accuracy of the estimation of the actual vehicle acceleration.

[0113] Below, referring to the vehicle control system shown in Figure 1, and taking a front-wheel-drive vehicle in a parking scenario as an example, we will explain the specific technical implementation of vehicle parking control using a combined lateral and longitudinal tracking control method.

[0114] First, the outer ring can be based on the longitudinal displacement s and longitudinal velocity v of the car. x And acceleration dynamics (acceleration a and desired acceleration a) d The longitudinal motion equations, and based on the vehicle's lateral position y and yaw angle... Lateral velocity v y yaw rate and front wheel steering angle dynamics (front wheel steering angle δ and front wheel steering angle control value δ) d The lateral motion equations are used to construct the parking trajectory model prediction equations. Then, by combining the lateral and longitudinal tracking target weights, the parking trajectory tracking control problem is transformed into a linear constraint optimization problem. The primordial dual interior point method can then be used to solve this linear constraint optimization problem to obtain the outer loop control quantity (front wheel steering angle control quantity δ). d and expected acceleration a d).

[0115] Specifically, the longitudinal motion state quantity of the car can be defined as X1 = [sv x a] T The longitudinal motion control variable of the car is U1 = a d Then the longitudinal motion equation of the car can be expressed as:

[0116] In the formula, This represents the derivative of X1, including the derivative of s (i.e., longitudinal velocity) and v. x The derivatives of a (i.e., acceleration) and a (i.e., jerk) are given, and A1 and B1 are expressed as follows:

[0117] In the formula, τ1 is the time constant of the inertial element, which characterizes the response characteristics of the vehicle's drive and braking system.

[0118] Referring to the vehicle control system shown in Figure 1, for the acceleration a in the longitudinal motion state quantity X1 of the vehicle, the acceleration a0 at the current moment can be the vehicle acceleration estimated in real time by the acceleration estimation module 103 based on the rolling time domain.

[0119] The lateral motion state variables of a car can be defined as follows: The lateral motion control variable of the car is U2 = δ d Furthermore, the lateral position of the car can be approximated as... The equation of motion for the car in the lateral direction can then be expressed as:

[0120] In the formula, A2 and B2 are respectively represented as:

[0121] In the formula, C f and C r For the lateral stiffness of the front and rear wheels of the car; l f and l r Let m and I be the distances from the car's center of gravity to the front and rear axles, respectively. z V represents the mass of the car and its moment of inertia about its vertical axis. x τ is the longitudinal velocity of the vehicle; τ2 is the time constant of the inertial element that characterizes the response of the vehicle's steering system.

[0122] Based on the above equations of longitudinal and lateral motion of the vehicle, the combined longitudinal and lateral state variables of the vehicle can be defined as follows: The combined lateral and longitudinal control quantity of the vehicle is U = [a d ,δ d ] T Then the combined lateral and longitudinal motion equations of the car can be expressed as:

[0123] In the formula, A and B are represented as follows:

[0124] Alternatively, the vehicle motion equation described in other embodiments of this application may be equation (3).

[0125] By discretizing equation (3) using the fourth-order Runge-Kutta integral method, the parking trajectory model prediction equation can be obtained to describe the lateral and longitudinal joint parking trajectory tracking control problem:

[0126] In the formula, the calculation step size h and coefficients K1, K2, K3 and K4 can be expressed as follows:

[0127] In the formula, t f M and M represent the prediction duration and the number of discrete points for model prediction control, respectively.

[0128] Referring to equations (4) to (5), the vehicle's state quantity X(k+1) at time k can be obtained based on the vehicle's state quantity X(k) and control quantity U(k) at time k, and this process can be repeated. In this way, the vehicle's state quantity and control quantity at a series of future time points can be predicted based on the planning information and the vehicle's current state quantity X0.

[0129] As shown in Figure 3, by considering the steering angle and longitudinal acceleration constraints of the front wheels of the vehicle, the lateral and longitudinal joint parking trajectory tracking control problem described by equation (4) can be transformed into the following constrained linear optimization problem:

[0130] In the formula, δ d max and a d max These represent the maximum steering angle of the front wheels and the maximum longitudinal acceleration of the vehicle, respectively; W1 and W2 are weighting coefficients; Z1 and Z2 can be expressed as:

[0131] Referring to equation (6), the linear optimization problem includes not only lateral tracking error (the error between the vehicle's lateral state variables and the corresponding planned reference values) but also longitudinal tracking error (the error between the vehicle's longitudinal state variables and the corresponding planned reference values) in order to achieve the purpose of joint lateral and longitudinal solution.

[0132] Practically, the first function described in other embodiments of this application can be the function in equation (6). The first weight and the second weight described in other embodiments of this application can be the weight coefficients W1 and W2 in equation (6), respectively.

[0133] Alternatively, the vehicle lateral control constraint condition described in other embodiments of this application can be -δ in equation (6). d max ≤δ d (k)≤δd max The longitudinal control constraint for the vehicle can be -a in equation (6). d max ≤a d (k)≤a d max .

[0134] The linear optimization problem described by equation (6) can be abstracted as follows:

[0135] Among them, the inequality constraints in equation (7) correspond to the two inequality constraints in equation (6), the equality constraints in equation (7) correspond to the two equality constraints in equation (6), and the minimization function in equation (7) corresponds to the minimization function in equation (6).

[0136] The following explanation, referring to equations (8) to (14), illustrates the process of using the primal dual interior point method to solve linear constrained optimization problems and obtain the outer loop control quantity.

[0137] By transforming the constraints in equation (7), we can obtain the first-order necessity condition for the original optimization problem shown in equation (7): -Ax+b=0 (8b) Cx-d≥0 (8c) μ(Cx-d)=0 (8d) μ≥0 (8e)

[0138] In the formula, L(x,λ,μ) is the Lagrangian function of the primal optimization problem, which can be expressed as:

[0139] In the formula, λ and μ are both Lagrange multipliers.

[0140] We can define the slack variable η = Cx - d ≥ 0, then the first-order necessity condition described by equation (8) can be transformed into: Hx + gA T λ-C T μ=0 (10a) -Ax+b=0 (10b) -Cx+d+η=0 (10c) μ i η i =0 (10d) (μ,η)≥0 (10e)

[0141] Introduce homotopy variable τ = μ i η i The first-order necessity condition described by relaxation formula (10) yields: Hx + gA T λ-C Tμ=0 (11a) -Ax+b=0 (11b) -Cx+d+η=0 (11c) UN-τI=0 (11d) (μ,η)≥0 (11e)

[0142] In the formula, U and N are diagonal matrices, and I is the identity matrix.

[0143] Based on equation (11), we can obtain:

[0144] In the formula, The Jacobian matrix can be represented as:

[0145] Further rearranging equation (12), we get:

[0146] In the formula, λ k+1 It can be represented as λ at the (k+1)th time point in the future (or at the (k+1)th step of the prediction).

[0147] Based on equation (14), the outer loop control quantity (front wheel steering angle control quantity δ) that satisfies the constraints of equation (14) and minimizes the function in equation (7) can be obtained. d and expected acceleration a d ).

[0148] The solution to the linear optimization problem can be obtained by iteratively solving equation (14), which yields the outer loop control quantity. Feasibly, based on the time sequence, the state quantity and control quantity at a series of future time points can be iterated sequentially from front to back, based on the current state quantity of the vehicle and the planned lateral and longitudinal reference values.

[0149] Referring to Figure 1, the vehicle can be controlled by the front wheel steering angle δ. d The steering mechanism of the vehicle body 100 is controlled to achieve lateral control of the vehicle. The inner ring can adjust the steering according to the desired acceleration a. d The vehicle's actual acceleration is estimated in real time, and an acceleration control quantity a is generated. Then, the drive braking mechanism of the vehicle body 100 is controlled according to the acceleration control quantity a to achieve the purpose of longitudinal vehicle control.

[0150] Thus, after obtaining the outer loop control quantity through the outer loop solution, the inner loop can employ a parallel operation of an unconstrained rolling time-domain estimation module with a time-domain length of 1 and a constrained rolling time-domain estimation module with a time-domain length of N to estimate the vehicle body acceleration that satisfies the physical constraints of the state variables (e.g., the vehicle's historical speed used to estimate vehicle acceleration meets the constraint on the magnitude of the longitudinal velocity). Subsequently, a PID controller is used to obtain the reference acceleration 'a' based on the outer loop solution. dGiven the estimated vehicle body acceleration a0, an acceleration control quantity a is output. By controlling the drive and braking mechanism of the vehicle body 100 using the acceleration control quantity a, precise tracking control of the planned parking speed can be achieved.

[0151] Among them, the unconstrained rolling time-domain estimation module with a time-domain length of 1 is used to iteratively calculate the arrival cost function (i.e., the output of the rolling time-domain estimation module with a time-domain length of 1), and the constrained rolling time-domain estimation module with a time-domain length of N estimates the vehicle body acceleration based on the arrival cost function and the real-time updated test data sequence period (i.e., the data sequence of the vehicle's longitudinal state quantities at the first N time points).

[0152] Specifically, the system state vector can be defined as x(t) = [vaj] T If the system noise is ω(t)~N(0,q), then the system state equation can be expressed as:

[0153] In the system state vector, v, a, and j represent the vehicle's longitudinal velocity, acceleration, and jerk, respectively. The system matrix F and the system noise covariance matrix q can be expressed as follows:

[0154] The system state equation shown in equation (15) can be discretized. Defining the system discretization period as Δt, from linear system theory, the discretized system state equation is obtained as: x k+1 =Ax k +ω k (16)

[0155] Wherein, the discretized system noise satisfies ω k ~N(0,Q), the discretized system matrix A and system noise covariance matrix Q can be expressed as:

[0156] According to Equation (16), the system state vector at time k+1 can be calculated based on the system state vector at time k and the system noise.

[0157] The system observation vector can be defined as y. k =[v k a k ] T If the measurement noise is υk~N(0,R), then the discretized system state equation can be rewritten as:

[0158] The measurement matrix C can be represented as:

[0159] In equation (17), vk It can represent the longitudinal velocity of the vehicle observed by the on-board sensor. The observed acceleration can be obtained by differentiating the observed longitudinal velocity. The measurement noise can be the noise of the on-board sensor.

[0160] Equation (17) defines the relationship between the system observation vector and measurement noise and the system state vector at the same time.

[0161] For the system state equation described by equation (17), a rolling time-domain estimator can be designed using a sequence of measurement data of length N in the time domain to estimate the state vector at the next moment in real time. According to the Bayes criterion, the conditional joint probability density function of the state vector sequence can be expressed as:

[0162] In Equation (18), the sequence y (i.e., the measurement data sequence) represents the system observation vector at each time point starting from time 0 (i.e., the time when the vehicle starts driving), and the sequence x (i.e., the state vector sequence) represents the system state vector at each time point starting from time 0. Equation (18) is used to describe the probability distribution of x given y, reflecting the probability distribution relationship between the system observation vector and the system state vector.

[0163] Assuming the measurement noise sequences are independent, the conditional joint probability density function of the measurement data sequences in equation (18) can be expressed as:

[0164] Wherein, the measurement probability p(y) k |x k The distribution is a Gaussian distribution with mean Cxk and covariance R, which can be expressed as:

[0165] Since the state vector sequence is a Markov chain, that is: state vector x k+1 Only related to the state vector x from the previous time step k and p(ω) k If related, then the joint probability density function of the state vector sequence in equation (18) can be expressed as:

[0166] Among them, the prior probability p(x0) of the initial state vector and the state vector transition probability p(x0) in equation (21) k+1 |x k This can be represented as:

[0167] Substituting equations (19) to (23) into equation (18), we get:

[0168] Given the sequence of measured data, the state vector sequence that maximizes the conditional joint probability density function described by equation (24) is the output of the rolling time-domain estimator. Therefore, the optimization problem concerning the state vector sequence to be solved can be established as follows:

[0169] That is, by solving equation (25), we can obtain the state vector sequence that maximizes the conditional joint probability density function described by equation (24), which can be used as the output of the rolling time domain estimator.

[0170] Considering the state equation constraints and state vector constraints of equation (17), the optimization problem described by equation (25) can be transformed into the following fully information-constrained optimization problem:

[0171] The inequality constraint in equation (26) is a constraint on the state quantity, which can make the solved state quantity conform to the set state quantity constraint.

[0172] The solution-finding process differs for T≤N and T>N. When T≤N, the number of observations is small, and the optimal solution can be obtained from all observations using a constrained rolling time domain of length N. When T>N, the number of observations is large, requiring a combination of an unconstrained rolling time domain of length 1 and a constrained rolling time domain of length N to obtain the optimal solution from all observations.

[0173] When T≤N, solve the fully information-constrained optimization problem described by equation (26) to obtain the optimal solution. Based on this optimized solution, the output of the constrained rolling time domain with a time domain length of N can be obtained.

[0174] Furthermore, based on the obtained optimized solution, the estimated value of the state vector can be obtained recursively using the state equation of equation (17):

[0175] When T > N, the full-information constrained optimization problem described by equation (26) is transformed into the following finite-dimensional constrained optimization problem using the arrival cost function:

[0176] Where, Θ(x) T-N The arrival cost function contains the remaining measurement data sequence. The impact on the estimated value can be obtained by solving an unconstrained rolling time-domain estimation problem with a time-domain length of 1. The arrival cost function is the output of the unconstrained rolling time-domain problem with a time-domain length of 1.

[0177] Solving the constrained optimization problem described by equation (28), we obtain the optimal solution. Based on this optimized solution, the output of the constrained rolling time domain with a time domain length of N can be obtained.

[0178] Furthermore, based on the obtained optimized solution, the estimated value of the state vector can be obtained recursively using the state equation of equation (17):

[0179] After obtaining the estimated value of the state vector, the estimated value of the acceleration can be obtained from it.

[0180] Below, referring to equations (30) to (35), for the arrival cost function Θ(x) in equation (28) T-N The solution process will be explained.

[0181] Referring to Figure 4, the state vector x is known. T-N-1 The prior estimates and covariance matrices are respectively and P T-N-1 Measurement data y can be used T-N-1 Real-time estimation of state vector x T-N Therefore, the unconstrained rolling time-domain estimation problem with a time-domain length of 1 is established as follows:

[0182] Where, Θ T-N-1 (x T-N-1 This can be represented as:

[0183] Where, σ T-N-1 With the state vector x to be solved T-N-1 and system noise ω T-N-1 Irrelevant.

[0184] Solving the unconstrained rolling time-domain estimation problem with a time-domain length of 1 described by equation (30), we can obtain the state vector x. T-N The estimated value is:

[0185] According to the definition of the arrival cost function, we can obtain:

[0186] Solving equation (32) yields the optimized value of the objective function:

[0187] Where, σ T-N With the state vector x to be solved T-N and system noise ω T-N Irrelevant, covariance matrix P T-N It can be represented as: P T-N =Q+A(P) T-N-1 -P T-N-1 C T (R+CP T-N-1 C T ) -1 CPT-N-1 A T (34)

[0188] The state vector x T-N The estimated value As the state vector x T-N The prior estimate, equation (33) can be modified as follows:

[0189] Therefore, by solving the unconstrained rolling time-domain estimation problem with a time-domain length of 1, the arrival cost function of the finite-dimensional constrained optimization problem described by equation (28) can be obtained.

[0190] Referring to Figure 4, the rolling time-domain estimation module with a time-domain length of 1 can obtain the system state variables, system observations, and covariance at time TN-1 based on the system state variables, system observations, and covariance at time TN-1. Referring to equation (28), the system state variables at time TN output by the rolling time-domain estimation module with a time-domain length of 1 can be reassigned to the system state variables at time TN used by the rolling time-domain estimation module with a time-domain length of N. The rolling time-domain estimation module with a time-domain length of N can then obtain the system state variables at time T-N+1 based on the system state variables at time TN, and can correct the system state variables at the corresponding time based on the system observations. This process is repeated until the system state variables of the vehicle at time T are obtained, thus obtaining the estimated actual acceleration value in the system state variables.

[0191] Based on the above, the basic steps for estimating the state vector (including acceleration) using the rolling time-domain algorithm can be summarized as follows: Step 1 to Step 6:

[0192] Step 1, initialize the prior estimate of the state vector as follows: The covariance matrix is ​​P0, the time domain length is set to N, and T = 1.

[0193] Step 2: Solve the unconstrained rolling time-domain estimation problem with a time-domain length of 1 described by equation (30) to obtain the state vector x. T-N The estimated value.

[0194] Step 3: Solving equation (32) yields the arrival cost function of the unconstrained rolling time-domain estimation problem described by equation (30), which can also be approximated as the arrival cost function of the constrained optimization problem described by equation (28).

[0195] Based on Steps 2 and 3, the arrival cost function can be calculated, which is the output of the unconstrained rolling time domain with a time domain length of 1. This output is used to calculate the optimization solution in Step 5 (i.e., the output of the constrained rolling time domain with a time domain length of N).

[0196] Step 4: When T≤N, solve the fully information-constrained optimization problem described by equation (26) to obtain the optimal solution. And the state vector x is obtained using equation (27). T The estimated value.

[0197] In Step 4, a constraint time domain of length N is used to roll the time domain to obtain the optimized solution.

[0198] Step 5: When T > N, solve the constrained optimization problem described by equation (28) to obtain the optimal solution. And the state vector x is obtained using equation (29). T The estimated value.

[0199] In Step 5, an unconstrained rolling time domain with a time domain length of 1 and a constrained rolling time domain with a time domain length of N are used to obtain the optimized solution.

[0200] Step 6, use the measured value y T Construct a new test data sequence, set T:=T+1, and return to Step 2.

[0201] After obtaining the optimized solution, an estimate of the vehicle's actual acceleration can be obtained. This process can then be repeated in a loop to estimate the vehicle's actual acceleration in real time.

[0202] As can be seen from the above, the lateral and longitudinal joint parking control method provided in this application embodiment can solve the problem of joint control of parking path and speed in scenarios with large parking path curvature change rate. Furthermore, by using the rolling time domain, the vehicle acceleration that satisfies the physical constraints of the state variables can be estimated, enabling the inner loop control to suppress the adverse effects of information transmission delay on parking speed tracking accuracy.

[0203] As shown in Figure 5, this application embodiment provides a vehicle control device 50, including: a first processing module 501, used to solve a tracking control problem for tracking a planned path and a planned speed based on the lateral motion state quantity and longitudinal motion state quantity of a first vehicle, and obtain the first wheel steering angle control quantity and the desired acceleration of the first vehicle; wherein, the first wheel is the front wheel or the rear wheel, the planned path includes reference values ​​of the lateral motion state quantity of the vehicle at a series of future time points, and the planned speed includes reference values ​​of the longitudinal motion state quantity of the vehicle at a series of future time points; a second processing module 502, used to obtain the acceleration control quantity of the first vehicle based on the desired acceleration and the actual acceleration of the first vehicle; and a control module 503, used to control the first vehicle based on the first wheel steering angle control quantity and the acceleration control quantity.

[0204] One embodiment of this application provides a chip, including: a processor, which is configured to execute computer program instructions stored in a memory, wherein when the computer program instructions are executed by the processor, the chip is triggered to execute the method described in any embodiment of this application.

[0205] One embodiment of this application provides an electronic device including at least one processor and a memory coupled together. The memory is used to store computer program instructions, and the processor is used to execute the computer program instructions. When the computer program instructions are executed by the processor, the electronic device is triggered to execute the method described in any embodiment of this application.

[0206] One embodiment of this application provides a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the methods described in any embodiment of this application.

[0207] One embodiment of this application provides a computer program product, which includes a computer program that, when run on a computer, causes the computer to perform the methods described in any embodiment of this application.

[0208] Figure 6 is a schematic diagram of a computer device provided in one embodiment of this application. As shown in Figure 6, the computer device 20 of this embodiment includes a processor 21 and a memory 22. The memory 22 is used to store a computer program 23 that can run on the processor 21. When the computer program 23 is executed by the processor 21, it implements the steps in the method embodiment of this application. To avoid repetition, these steps are not described in detail here. Alternatively, when the computer program 23 is executed by the processor 21, it implements the functions of each model / unit in the device embodiment of this application. To avoid repetition, these functions are not described in detail here.

[0209] Computer device 20 includes, but is not limited to, processor 21 and memory 22. Those skilled in the art will understand that FIG6 is merely an example of computer device 20 and does not constitute a limitation on computer device 20. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0210] The processor 21 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or it can be any conventional processor.

[0211] The memory 22 can be an internal storage unit of the computer device 20, such as a hard disk or RAM of the computer device 20. The memory 22 can also be an external storage device of the computer device 20, such as a plug-in hard disk, Smart Media (SM) card, Secure Digital (SD) card, or FlashCard equipped on the computer device 20. Furthermore, the memory 22 can include both internal and external storage units of the computer device 20. The memory 22 is used to store the computer program 23 and other programs and data required by the computer device. The memory 22 can also be used to temporarily store data that has been output or will be output.

[0212] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, apparatuses, or units, and may be electrical, mechanical, or other forms.

[0213] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0214] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units.

[0215] An integrated unit implemented as a software functional unit can be stored in a computer-readable storage medium. This software functional unit, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk.

[0216] In this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0217] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0218] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments of this application can be implemented using electronic hardware, computer software, or a combination of electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0219] Those skilled in the art will readily understand that, for the sake of convenience and brevity, the same or similar parts between the various embodiments of this application can be referred to mutually. For example, the specific working processes of the systems, devices, and units described in the embodiments of this application can be referred to the corresponding processes in the method embodiments of this application, and will not be repeated here.

[0220] The above description is merely a specific embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A vehicle control method, characterized in that, include: Based on the lateral and longitudinal motion state variables of the first vehicle, solve the tracking control problem for tracking the planned path and speed, and obtain the first wheel steering angle control variable and the desired acceleration of the first vehicle. Wherein, the first wheel is either the front wheel or the rear wheel, the planned path includes reference values ​​of the vehicle's lateral motion state at a series of future time points, and the planned speed includes reference values ​​of the vehicle's longitudinal motion state at a series of future time points. Based on the desired acceleration of the first vehicle and the actual acceleration of the first vehicle, the acceleration control amount of the first vehicle is obtained; The first vehicle is controlled based on the first wheel steering angle control amount and acceleration control amount.

2. The method according to claim 1, characterized in that, The tracking control problem includes the vehicle motion equations; The state variables of the vehicle motion equation include at least one of the following: lateral position, heading angle, yaw angle, lateral velocity, heading angular velocity, yaw angular velocity, and first wheel steering angle; and at least one of the following: longitudinal displacement, longitudinal velocity, and acceleration. The control variables in the vehicle motion equation include: the first wheel steering angle control variable and the desired acceleration.

3. The method according to claim 1, characterized in that, The tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimize the first function; The first function includes path tracking error and speed tracking error, wherein the path tracking error is obtained based on the error between the vehicle's lateral motion state quantity and the reference value in the planned path, and the speed tracking error is obtained based on the error between the vehicle's longitudinal motion state quantity and the reference value in the planned speed.

4. The method according to claim 3, characterized in that, The tracking control problem is used to solve for the first wheel steering angle control quantity and the desired acceleration that minimize the first function; The first function includes a process constraint control quantity and a first weight of the process constraint control quantity, a terminal constraint control quantity and a second weight of the terminal constraint control quantity; The first weight and the second weight are obtained based on the scenario in which the first vehicle is located; The terminal constraint control parameters include the path tracking error and speed tracking error at the future target time point; The process constraint control quantities include path tracking error and speed tracking error at each time point within the future time period, and the target time point is the next time point of the future time period.

5. The method according to claim 3, characterized in that, The tracking control problem is a linear constraint optimization problem that includes the first function, and the constraints of the linear constraint optimization problem include vehicle lateral control constraints and vehicle longitudinal control constraints. The solution to the tracking control problem, which involves tracking the planned path and planned speed, includes: The linearly constrained optimization problem is solved using the primal dual interior point method.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: The actual acceleration of the first vehicle is estimated using a rolling time-domain estimation module with a time-domain length of 1 and a rolling time-domain estimation module with a time-domain length of N, where N > 1; The rolling time-domain estimation module with a time-domain length of 1 is used to obtain the first parameter based on the measured value of the longitudinal speed of the first vehicle in the first time period when the number of time points of the vehicle's driving time is greater than N. The rolling time-domain estimation module with a time-domain length of N is used to estimate the actual acceleration of the first vehicle based on the first parameter and the measured value of the longitudinal velocity of the first vehicle in the second time period when the number of time points of the vehicle's driving time is greater than N. The second time period is connected to the first time period, and the starting point of the first time period is the initial travel time of the first vehicle.

7. The method according to any one of claims 1-5, characterized in that, The step of obtaining the acceleration control amount of the first vehicle based on the desired acceleration and the actual acceleration of the first vehicle includes: The desired acceleration of the first vehicle and the actual acceleration of the first vehicle are input into the proportional-integral-derivative (PID) controller to obtain the acceleration control quantity of the first vehicle output by the PID controller.

8. A vehicle control device, characterized in that, include: The first processing module is used to solve the tracking control problem for tracking the planned path and planned speed based on the lateral motion state quantity and longitudinal motion state quantity of the first vehicle, and to obtain the first wheel steering angle control quantity and the desired acceleration of the first vehicle; wherein, the first wheel is the front wheel or the rear wheel, the planned path includes reference values ​​of the vehicle's lateral motion state quantity at a series of future time points, and the planned speed includes reference values ​​of the vehicle's longitudinal motion state quantity at a series of future time points. The second processing module is used to obtain the acceleration control amount of the first vehicle based on the expected acceleration of the first vehicle and the actual acceleration of the first vehicle. The control module is used to control the first vehicle based on the first wheel steering angle control amount and acceleration control amount of the first vehicle.

9. An electronic device, characterized in that, The electronic device includes at least one processor coupled to a memory for storing computer program instructions, and the processor for executing the computer program instructions, wherein when the computer program instructions are executed by the processor, the electronic device is triggered to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when run on a computer, causes the computer to perform the method as described in any one of claims 1-7.

Images