Integrated model prediction obstacle avoidance planning and control method

Abstract

The invention discloses an integrated model prediction obstacle avoidance planning and control method. The method comprises an obstacle avoidance behavior guide track generation layer and an obstacle avoidance track planning and control layer. The obstacle avoidance behavior guide track generation layer comprehensively considers the influence of the vehicle speed, the road surface state and the obstacle size to obtain a guide track, and determines the expected values of the longitudinal speed, the lateral position and the yaw angle in the vehicle obstacle avoidance process; the obstacle avoidance trajectory planning and control layer comprises a prediction model, an objective function, constraint conditions and an optimization solution link; the objective function comprises an obstacle avoidance behavior guide trajectory tracking precision item, a control quantity item and a control quantity increment item; the constraint conditions comprise obstacle avoidance constraint, control quantity constraint and control quantity increment constraint; the control quantity comprises a front wheel turning angle and wheel longitudinal force, and the wheel longitudinal force comprises a left front wheel longitudinal force, a right front wheel longitudinal force, a left rear wheel longitudinal force and a right rear wheel longitudinal force; obstacle avoidance control with high safety and good calculation real-time performance is realized under different driving conditions.

Description

Technical Field

[0001] The present invention relates to the field of vehicle obstacle avoidance control, and particularly to an integrated model predictive obstacle avoidance trajectory planning and control method. Background Art

[0002] In current production and life, assisted driving and autonomous driving have gradually integrated into life; it is particularly important to achieve safe and real-time obstacle avoidance control during the autonomous driving process.

[0003] The model predictive control algorithm (MPC) has the advantage of being able to effectively handle multi-constraint and multi-objective optimization problems and has become one of the mainstream algorithms for obstacle avoidance planning and control; according to the coupling degree of the trajectory planning function and the trajectory tracking function, the autonomous driving obstacle avoidance planning and control framework based on MPC can be divided into a hierarchical obstacle avoidance framework and an integrated obstacle avoidance framework; compared with the hierarchical obstacle avoidance framework, the integrated obstacle avoidance framework comprehensively considers multi-objective requirements and constraints such as obstacle avoidance and execution under a unified optimization problem framework to solve the optimal control sequence, effectively reducing the obstacle avoidance planning and control coordination problem and achieving good obstacle avoidance control effects. However, the feasible solution space of the traditional integrated obstacle avoidance planning and control framework is broad, the solution time is long, and it is difficult to apply in real time; therefore, it is necessary to further optimize the integrated model predictive obstacle avoidance planning and control framework to improve the balance effect of obstacle avoidance safety and operation real-time performance. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides an integrated model predictive obstacle avoidance planning and control method, which establishes an obstacle avoidance behavior guiding trajectory by comprehensively considering the influence of vehicle speed, road surface conditions, and obstacle size, and constructs an integrated model predictive obstacle avoidance planning and control algorithm on this basis to comprehensively improve obstacle avoidance safety and operation real-time performance.

[0005] According to the present invention, an integrated model predictive obstacle avoidance planning and control method is provided, including an obstacle avoidance behavior guiding trajectory generation layer and an obstacle avoidance trajectory planning and control layer; The guiding trajectory is obtained by the obstacle avoidance behavior guiding trajectory generation layer by comprehensively considering the influence of vehicle speed, road surface conditions, and obstacle size, and the expected values of the longitudinal speed, lateral position, and yaw angle during the vehicle obstacle avoidance process are determined; The obstacle avoidance trajectory planning and control layer establishes an obstacle avoidance optimization problem based on the model predictive control framework, including a prediction model, an objective function and constraint conditions, and an optimization solution link; The prediction model is established based on a three-degree-of-freedom vehicle dynamics model; the prediction model obtains the vehicle state change within the prediction time domain according to the current moment state and control input, and assists in determining the optimal control quantity in combination with the objective function and constraint conditions; The objective function includes an item for the tracking accuracy of the obstacle avoidance behavior guiding trajectory, an item for the control quantity, and an item for the increment of the control quantity. The objective function quantifies and characterizes the obstacle avoidance safety and control smoothness of the vehicle by softly constraining the vehicle state, control quantity, and control quantity increment within the prediction time domain. The control quantity corresponding to the minimum value of the objective function is the optimal control quantity; The vehicle state within the prediction time domain includes the longitudinal speed, lateral position, and yaw angle, and the obstacle avoidance safety is improved by the expected values obtained from the approaching obstacle avoidance behavior guiding trajectory; The constraint conditions include obstacle avoidance constraints, control quantity constraints, and control quantity increment constraints. The control quantities that meet the constraint conditions are selected as candidates for the optimized control quantity through the constraint conditions to enhance the obstacle avoidance safety and control smoothness; The optimization solution uses a sequential quadratic programming solver and sets iterative warm start to solve and obtain the optimized control sequence within the control time domain, and takes the first item of it as the optimized control quantity for actual output; The control quantity includes the front wheel steering angle and the wheel longitudinal force, and the wheel longitudinal force includes the left front wheel longitudinal force, the right front wheel longitudinal force, the left rear wheel longitudinal force, and the right rear wheel longitudinal force.

[0006] Compared with the prior art, the present invention has the following beneficial effects: The obstacle avoidance optimization problem is divided into an obstacle avoidance behavior guiding trajectory generation layer and an obstacle avoidance trajectory planning and control layer; the obstacle avoidance behavior guiding trajectory generation layer generates a rough guiding trajectory based on a fifth-degree polynomial, which provides a reference for the model predictive optimization solution, helps to focus the optimization iterative solution near the feasible guiding trajectory to improve the optimization solution efficiency; the obstacle avoidance trajectory planning and control layer designs an integrated model predictive control algorithm on the basis of the guiding trajectory, further considering constraints such as vehicle dynamics, obstacle avoidance smoothness, and obstacle avoidance safety to enhance the obstacle avoidance safety, and solves to obtain a matching optimized control sequence; this method makes full use of the high efficiency of polynomial curve obstacle avoidance trajectory planning and the multi-constraint and multi-objective optimization control ability of model predictive control, and improves the calculation efficiency while enhancing the obstacle avoidance safety and control smoothness through iterative optimization solution steps; compared with the traditional integrated obstacle avoidance framework, it better balances the obstacle avoidance safety and operation real-time performance, and alleviates the problem that it is difficult to apply in real time due to the wide feasible solution space of the traditional integrated obstacle avoidance framework.

[0007] Furthermore, obtain the vehicle longitudinal speed , dimension information, and safety threshold information; the dimension information includes the vehicle width of the own vehicle , the distance from the center of mass of the own vehicle to the front of the vehicle , and the width of the obstacle ; The safety threshold information includes the maximum value of the lateral acceleration ; At the vehicle longitudinal speed , establish the vehicle obstacle avoidance behavior guidance trajectory formula through a fifth-degree polynomial based on size information; Obtain the obstacle avoidance behavior guidance trajectory based on the vehicle obstacle avoidance behavior guidance trajectory formula in combination with safety threshold information ; Combine the vehicle obstacle avoidance trajectory information to obtain the vehicle's desired yaw angle through Formula 1 .

[0008] The beneficial effect of the previous step is that by comprehensively considering the influence of vehicle state information, obstacle size, and road adhesion conditions on obstacle avoidance behavior, the obstacle avoidance behavior guidance trajectory can be flexibly generated under different driving conditions, providing a more effective reference for subsequent obstacle avoidance planning and control; Considering obstacle avoidance safety and comfort, divide the safety threshold information into stable obstacle avoidance safety threshold information and emergency obstacle avoidance safety threshold information, and substitute them into the obstacle avoidance behavior guidance trajectory formula to obtain the stable obstacle avoidance behavior guidance trajectory and the emergency obstacle avoidance behavior guidance trajectory respectively.

[0009] Furthermore, the vehicle obstacle avoidance behavior guidance trajectory formula is: ; where, , , , , , ; ; ; , , , , , are the fitting coefficients of the obstacle avoidance trajectory curve; is the lateral displacement of obstacle avoidance; is the lateral reserved safety distance.

[0010] The beneficial effect of the previous step is that establishing the obstacle avoidance behavior guidance trajectory formula based on a fifth-degree polynomial, on the one hand, the number of quantities to be solved is greatly reduced, the solution difficulty is reduced, and the calculation efficiency is improved. On the other hand, the fitting coefficient calculation formula comprehensively considers factors such as vehicle speed, obstacle size, and road surface adhesion conditions, which is beneficial to adapting to vehicle speed changes and flexibly generating matching obstacle avoidance behavior guidance trajectories under the influence of obstacles of different sizes.

[0011] Furthermore, the vehicle obstacle avoidance behavior guidance trajectory is: ; ; The The longitudinal position of the vehicle's center of mass at the moment during the obstacle avoidance process, and the lateral position of the vehicle's center of mass at the moment during the obstacle avoidance process; The obstacle avoidance time is calculated as follows: {t}_{f}=\sqrt[{3}] {\frac {10\sqrt {3}\ast {Y}_{w}} {3\ast {a}_{ymax}}} ; where, ; .

[0012] Furthermore, by combining the obstacle avoidance behavior guiding trajectory with Formula 1, the desired yaw angle of the vehicle is obtained. The Formula 1 is: ; ; ; is the sampling moment; is the change in lateral position between adjacent moments; is the change in longitudinal position between adjacent moments.

[0013] The advantage of the previous step is to obtain the desired yaw angle of the vehicle.

[0014] Furthermore, the maximum lateral acceleration includes the maximum lateral acceleration for smooth obstacle avoidance and the maximum lateral acceleration for emergency obstacle avoidance; The maximum lateral acceleration for smooth obstacle avoidance is ; The maximum lateral acceleration for emergency obstacle avoidance is ; The is the road adhesion coefficient; The is the gravitational acceleration; The autonomous driving vehicle autonomously switches to smooth obstacle avoidance behavior or emergency obstacle avoidance behavior according to the driver's personal habits and actual obstacle avoidance conditions; When the smooth obstacle avoidance behavior is selected, is the maximum lateral acceleration for smooth obstacle avoidance, and the specific value is , and substituting it into the vehicle obstacle avoidance behavior guiding trajectory formula to obtain the smooth obstacle avoidance behavior guiding trajectory; When the emergency obstacle avoidance behavior is selected, is the maximum lateral acceleration for emergency obstacle avoidance, and the specific value is , substituting into the vehicle obstacle avoidance behavior guidance trajectory formula to obtain the emergency obstacle avoidance behavior guidance trajectory.

[0015] The advantage of adopting the previous step is that considering the requirements of obstacle avoidance safety and comfort, the smooth obstacle avoidance behavior guidance trajectory and the emergency obstacle avoidance behavior guidance trajectory are set, and the autonomous vehicle can select a reasonable obstacle avoidance behavior based on the actual obstacle avoidance conditions.

[0016] Furthermore, the prediction model includes the following formulas: ; ; where ; ; ; ; is the front wheel steering angle, , , , are the longitudinal forces of the four wheels, , , , are the lateral forces of the four wheels, is the longitudinal resistance, is the additional yaw moment, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle of the vehicle, is the vehicle's track width, is the moment of inertia of the vehicle about the z-axis, is the vehicle mass, is the vehicle's longitudinal speed, is the vehicle's lateral speed, is the vehicle's yaw angle, is the vehicle's yaw angular velocity, is the longitudinal acceleration rate, is the lateral acceleration rate, is the longitudinal speed in the geodetic coordinate, is the lateral speed in the geodetic coordinate.

[0017] Furthermore, the objective function is: ; is the sampling time; is the prediction horizon; is the control horizon; the first term of the objective function is the obstacle avoidance behavior guidance trajectory tracking accuracy term; in the first term of the objective function is the predicted value of the vehicle state information within the prediction horizon. In the first term of the objective function, is the expected value of the vehicle state information within the prediction horizon determined based on the obstacle avoidance behavior guiding trajectory; the second term is the control quantity term; the third term is the control quantity increment term; and and are the weight matrices corresponding to the three terms; is the weight matrix corresponding to the obstacle avoidance behavior guiding trajectory tracking accuracy term. The weight matrix includes the longitudinal speed tracking deviation weight, the vehicle lateral position tracking deviation weight, and the vehicle yaw angle tracking deviation weight; the longitudinal speed tracking deviation weight value is 2000 - 5000; the lateral position tracking deviation weight value is 5000 - 8000; the vehicle yaw angle tracking deviation weight value is 5000 - 8000; is the weight matrix corresponding to the control quantity term. The weight matrix includes the front wheel steering angle weight and the wheel longitudinal force weight; the front wheel steering angle weight value is 50000 - 80000; the wheel longitudinal force weight value is 10000 - 20000; is the weight matrix corresponding to the control quantity increment term. The weight matrix includes the front wheel steering angle increment weight and the wheel longitudinal force increment weight; the front wheel steering angle increment weight value is 50000 - 80000; the wheel longitudinal force increment weight value is 10000 - 20000.

[0018] Furthermore, the control quantity constraint conditions include: ; the control quantity increment constraint conditions include: ; The control quantity mainly includes the front wheel steering angle and the longitudinal forces of the four wheels , , , ; , are the maximum and minimum values of the control quantity respectively, determined by the physical limit boundary of the actuator; , are the maximum and minimum values of the control quantity increment respectively; The obstacle avoidance constraint requires that the vehicle always be within the drivable area, that is, the four vertices of the vehicle are within the safety boundary range to ensure the safety of vehicle obstacle avoidance; the constraint condition expression is: ; ; ; ; , , , ; , , , are the predicted values of the lateral positions of the four vertices of the vehicle, is the lateral position of the vehicle's center of mass, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle, is the vehicle's track width, is the vehicle's yaw angle; , , , , , , , are respectively the obstacle avoidance safety boundary values corresponding to the four vertices of the vehicle.

[0019] The advantage of adopting the previous step is that based on the vehicle state values at the current moment, the vehicle state changes within the prediction time domain for the longitudinal speed, yaw angular velocity, and vehicle position are obtained through the prediction model. Combining the vehicle physical constraints such as longitudinal force and steering angle and the dynamic limitations such as lateral acceleration and tire adhesion conditions to ensure the feasibility of the solved control strategy, and considering multi-objective requirements such as obstacle avoidance safety and control smoothness to construct an objective function to comprehensively evaluate the control sequence, and assist in solving the optimized obstacle avoidance control sequence. Inputting the optimized control quantity into the controlled vehicle, and repeating this process to form a rolling optimization mechanism, which can improve the adaptability of obstacle avoidance planning and control to the dynamic environment; For the objective function part, it mainly includes the trajectory tracking accuracy term for obstacle avoidance behavior guidance, the control quantity term, and the control quantity increment term. The first term is mainly to make full use of the guidance trajectory to avoid obstacles as much as possible; the second term is mainly to minimize the energy loss of the vehicle; the third term requires the vehicle control to be as smooth as possible.

[0020] For the constraint condition part, it mainly includes obstacle avoidance constraints, control quantity constraints, and control quantity increment constraints. The obstacle avoidance constraints generate an obstacle avoidance safety corridor based on the obstacle information and road boundary information to construct the vehicle vertex position boundary constraints, so as to ensure that the vehicle is always within the feasible range, which helps to enhance the obstacle avoidance safety; the control quantity and control quantity increment constraints reflect the requirements for economy and control smoothness.

[0021] Furthermore, use sequential quadratic programming to solve the integrated nonlinear optimization problem, and move the obtained optimized solution sequence one step forward as the initial solution for the next time step , thus forming an iterative warm start; {U}^{\ast}(t - 1)=\left [ {{u}^{\ast}_{t - 1|t - 1}, {u}^{\ast}_{t|t - 1}, {u}^{\ast}_{t + 1|t - 1}, \cdots \cdots, {u}^{\ast}_{t + {N}_{c} - 2|t - 1}} \right ]^{T} ; U{(t)}_{initial}=\left [ {{u}^{\ast}_{t|t - 1}, {u}^{\ast}_{t + 1|t - 1}, {u}^{\ast}_{t + 2|t - 1}, \cdots \cdots, {u}^{\ast}_{t + {N}_{c} - 2|t - 1}} \right ]^{T} .

[0022] The advantage of adopting the previous step is that the warm start setting makes full use of the solution result of the previous moment to enhance the real - time performance of the operation, and assists in quickly solving the optimized obstacle - avoidance control sequence, including the front - wheel steering angle and the longitudinal forces of the four wheels. Brief Description of the Drawings

[0023] Figure 1 It is a comparison diagram of the obstacle - avoidance trajectories between the obstacle - avoidance trajectory planning and control method proposed by the present invention and the traditional integrated model - predictive obstacle - avoidance planning and control method; Figure 2 It is a comparison diagram of the obstacle - avoidance trajectory curvatures between the obstacle - avoidance trajectory planning and control method proposed by the present invention and the traditional integrated model - predictive obstacle - avoidance planning and control method; Figure 3 It is a comparison diagram of the longitudinal speeds between the obstacle - avoidance trajectory planning and control method proposed by the present invention and the traditional integrated model - predictive obstacle - avoidance planning and control method; Figure 4 It is a comparison diagram of the front - wheel steering angle increments between the obstacle - avoidance trajectory planning and control method proposed by the present invention and the traditional integrated model - predictive obstacle - avoidance planning and control method; Figure 5 It is a comparison diagram of the total longitudinal force increments between the obstacle - avoidance trajectory planning and control method proposed by the present invention and the traditional integrated model - predictive obstacle - avoidance planning and control method. Detailed Embodiment

[0024] In order to better understand the technical solution of the present invention, the present invention will be further described below in conjunction with specific embodiments and the accompanying drawings of the specification. Embodiment 1:

[0025] According to this embodiment, an integrated model predictive obstacle avoidance planning and control method is provided, including an obstacle avoidance behavior guiding trajectory generation layer and an obstacle avoidance trajectory planning and control layer; The guiding trajectory is obtained by the obstacle avoidance behavior guiding trajectory generation layer considering the influences of vehicle speed, road surface condition, and obstacle size comprehensively, and the expected values of the longitudinal speed, lateral position, and yaw angle during the vehicle obstacle avoidance process are determined; The generating steps of the obstacle avoidance behavior guiding trajectory are as follows: Obtain the vehicle longitudinal speed , size information, and safety threshold information; the size information includes the vehicle width of the own vehicle , the distance from the centroid of the own vehicle to the front of the vehicle , and the width of the obstacle ; The safety threshold information includes the maximum value of the lateral acceleration ; Based on the vehicle longitudinal speed and the size information, establish the vehicle obstacle avoidance behavior guiding trajectory formula by a fifth-degree polynomial; Based on the vehicle obstacle avoidance behavior guiding trajectory formula and combined with the safety threshold information, obtain the obstacle avoidance behavior guiding trajectory ; Combine the vehicle obstacle avoidance trajectory information and obtain the expected yaw angle of the vehicle through Formula 1 .

[0026] The vehicle obstacle avoidance behavior guiding trajectory formula is: ; where, , , , , , ; ; ; , , , , , are the fitting coefficients of the obstacle avoidance trajectory curve; is the obstacle avoidance lateral displacement; is the lateral reserved safety distance.

[0027] The vehicle obstacle avoidance behavior guiding trajectory is: ; ; The is the centroid of the vehicle during the obstacle avoidance process at Longitudinal position at a moment is the lateral position of the vehicle's center of mass during the obstacle avoidance process at a moment; The obstacle avoidance time is calculated as: {t}_{f}=\sqrt[{3}] {\frac {10\sqrt {3}\ast {Y}_{w}} {3\ast {a}_{ymax}}} ; where, ; .

[0028] The vehicle's desired yaw angle is obtained by combining the obstacle avoidance behavior guidance trajectory with Formula 1. The said Formula 1 is: ; ; ; ; is the sampling moment; is the change in lateral position between adjacent moments; is the change in longitudinal position between adjacent moments.

[0029] The obstacle avoidance trajectory planning and control layer establishes an obstacle avoidance optimization problem based on the model predictive control framework, including a prediction model, an objective function, constraint conditions, and an optimization solution link; The prediction model is established based on a three-degree-of-freedom vehicle dynamics model; the prediction model can obtain the vehicle state changes within the prediction time domain according to the current moment state and control input, and assist in determining the optimal control quantity in combination with the objective function and constraint conditions; The prediction model includes the following formulas: ; ; where, ; ; ; ; is the front wheel steering angle, , , , are the longitudinal forces of the four wheels, , , , are the lateral forces of the four wheels, is the longitudinal resistance is the additional yaw moment, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle of the vehicle, is the vehicle's track width, is the moment of inertia of the vehicle about the z-axis, is the vehicle mass, is the longitudinal speed of the vehicle, is the lateral speed of the vehicle, is the yaw angle of the vehicle, is the yaw angular velocity of the vehicle, is the longitudinal speed change rate, is the lateral speed change rate, is the longitudinal speed in geodetic coordinates, is the lateral speed in geodetic coordinates.

[0030] The objective function includes an obstacle avoidance behavior-guided trajectory tracking term, a control quantity term, and a control quantity increment term. The objective function quantifies and characterizes the obstacle avoidance safety and control smoothness of the vehicle by imposing soft constraints on the vehicle state, control quantity, and control quantity increment within the prediction horizon. The control quantity corresponding to the minimum value of the objective function is the optimal control quantity; The objective function is: ; is the sampling time; is the prediction horizon; is the control horizon; The first term of the objective function is the obstacle avoidance behavior-guided trajectory tracking accuracy term; In the first term of the objective function is the predicted value of the vehicle state information within the prediction horizon, and in the first term of the objective function is the expected value of the vehicle state information within the prediction horizon determined based on the obstacle avoidance behavior-guided trajectory; The second term is the control quantity term; The third term is the control quantity increment term; , , are the weight matrices corresponding to the three terms; is the weight matrix corresponding to the obstacle avoidance behavior-guided trajectory tracking accuracy term, and the weight matrix includes the longitudinal speed tracking deviation weight, the vehicle lateral position tracking deviation weight, and the vehicle yaw angle tracking deviation weight; The longitudinal speed tracking deviation weight value is 3500; The lateral position tracking deviation weight value is 6500; The vehicle yaw angle tracking deviation weight value is 6500; is the weight matrix corresponding to the control quantity term, and the weight matrix It includes the front wheel steering angle weight and the wheel longitudinal force weight; the front wheel steering angle weight value is 65000; the wheel longitudinal force weight value is 15000; is the weight matrix corresponding to the control quantity increment term, and the weight matrix includes the front wheel steering angle increment weight and the wheel longitudinal force increment weight; the front wheel steering angle increment weight value is 65000; the wheel longitudinal force increment weight value is 15000.

[0031] The constraint conditions include obstacle avoidance constraints, control quantity constraints, and control quantity increment constraints. The optimal control quantity is screened out through the constraint conditions to enhance obstacle avoidance safety and control smoothness; The control quantity constraint conditions include: ; The control quantity increment constraint conditions include: ; The control quantity mainly includes the front wheel steering angle and the longitudinal forces of the four wheels , , , ; , are the maximum and minimum values of the control quantity respectively, which are determined by the physical limit boundary of the actuator; , are the maximum and minimum values of the control quantity increment respectively; The obstacle avoidance constraint requires that the vehicle is always within the drivable area, that is, the four vertices of the vehicle are within the safety boundary range to ensure the obstacle avoidance safety of the vehicle; the constraint condition expression is: ; ; ; ; , , , ; , , , are the predicted lateral position values of the four vertices of the vehicle, is the lateral position of the vehicle's center of mass, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle, is the vehicle's track width, is the vehicle's yaw angle; , , , , , , , They are the obstacle avoidance safety boundary values corresponding to the four vertices of the vehicle respectively.

[0032] The optimization solution model uses a sequential quadratic programming solver and sets iterative warm start. The optimized control sequence within the control time domain is obtained by solving, and the first item of it is used as the optimized control quantity of the actual output. Use sequential quadratic programming to solve the integrated nonlinear optimization problem, and use the obtained optimized solution sequence After moving one step forward as the initial solution for the next time step , thus forming an iterative warm start. {U}^{\ast}(t - 1)=\left [ {{u}^{\ast}_{t - 1|t - 1}, {u}^{\ast}_{t|t - 1}, {u}^{\ast}_{t + 1|t - 1}, \cdots \cdots, {u}^{\ast}_{t + {N}_{c} - 2|t - 1}} \right ]^{T} ; U{(t)}_{initial}=\left [ {{u}^{\ast}_{t|t - 1}, {u}^{\ast}_{t + 1|t - 1}, {u}^{\ast}_{t + 2|t - 1}, \cdots \cdots, {u}^{\ast}_{t + {N}_{c} - 2|t - 1}} \right ]^{T} .

[0033] The control quantity includes the front wheel steering angle and the wheel longitudinal force. The wheel longitudinal force includes the left front wheel longitudinal force, the right front wheel longitudinal force, the left rear wheel longitudinal force, and the right rear wheel longitudinal force.

[0034] Embodiment 2: The content that is the same as that in Embodiment 1 will not be elaborated here; the different solutions from Embodiment 1 are as follows: According to the integrated model predictive obstacle avoidance planning and control method provided in this embodiment, the following steps are further included: The maximum lateral acceleration includes the maximum lateral acceleration for smooth obstacle avoidance and the maximum lateral acceleration for emergency obstacle avoidance; The maximum lateral acceleration for smooth obstacle avoidance is ; The maximum lateral acceleration for emergency obstacle avoidance is ; The is the road adhesion coefficient; the is the gravitational acceleration; The autonomous vehicle autonomously switches to a smooth obstacle avoidance behavior or an emergency obstacle avoidance behavior according to the driver's personal habits and actual obstacle avoidance conditions; When the smooth obstacle avoidance behavior is selected, is the maximum value of the lateral acceleration for smooth obstacle avoidance, and the specific value is , which is substituted into the vehicle obstacle avoidance behavior guidance trajectory formula to obtain the smooth obstacle avoidance behavior guidance trajectory; When the emergency obstacle avoidance behavior is selected, is the maximum value of the lateral acceleration for emergency obstacle avoidance, and the specific value is , which is substituted into the vehicle obstacle avoidance behavior guidance trajectory formula to obtain the emergency obstacle avoidance behavior guidance trajectory.

[0035] The comparison table 1 of the calculation efficiency between the integrated model predictive obstacle avoidance planning and control method proposed in Embodiment 1 and the traditional integrated obstacle avoidance framework planning and control method is as follows: Table 1 Obstacle avoidance control algorithm Time-consuming for generating the obstacle avoidance behavior guidance trajectory (s) Time-consuming for obstacle avoidance planning and control (s) Total time-consuming (s) Traditional integrated obstacle avoidance algorithm —— 21.9946 21.9946 Example 1 0.20041 1.8558 2.0562 .

[0036] The above description is only a preferred embodiment of the present application and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features have similar functions to those disclosed in the present application (but not limited to).

Claims

1. An integrated model predictive obstacle avoidance planning and control method, characterized in that it includes an obstacle avoidance behavior guiding trajectory generation layer and an obstacle avoidance trajectory planning and control layer; the obstacle avoidance behavior guiding trajectory generation layer comprehensively considers the influences of vehicle speed, road surface conditions, and obstacle size to obtain a guiding trajectory, and determines the expected values of the longitudinal speed, lateral position, and yaw angle during the vehicle obstacle avoidance process; the obstacle avoidance trajectory planning and control layer establishes an obstacle avoidance optimization problem based on the model predictive control framework, including a prediction model, an objective function and constraint conditions, and an optimization solution link; the prediction model is established based on a three-degree-of-freedom vehicle dynamics model; the prediction model obtains the vehicle state changes within the prediction time domain according to the current moment state and control input, and combines the objective function and constraint conditions to assist in determining the optimal control quantity; the objective function includes an obstacle avoidance behavior guiding trajectory tracking accuracy term, a control quantity term, and a control quantity increment term. The objective function quantifies and characterizes the vehicle obstacle avoidance safety and control smoothness by softly constraining the vehicle state, control quantity, and control quantity increment within the prediction time domain. The control quantity corresponding to the minimum value of the objective function is the optimal control quantity; the vehicle state within the prediction time domain includes longitudinal speed, lateral position, and yaw angle, and the expected value obtained by approaching the obstacle avoidance behavior guiding trajectory is used to improve the obstacle avoidance safety; the constraint conditions include obstacle avoidance constraints, control quantity constraints, and control quantity increment constraints. The control quantities that meet the constraint conditions are selected as candidate items for the optimal control quantity through the constraint conditions to enhance the obstacle avoidance safety and control smoothness; the optimization solution uses a sequential quadratic programming solver and sets iterative warm start to solve and obtain the optimal control sequence within the control time domain, and takes the first item of it as the actually output optimal control quantity; the control quantity includes the front wheel steering angle and the wheel longitudinal force, and the wheel longitudinal force includes the left front wheel longitudinal force, the right front wheel longitudinal force, the left rear wheel longitudinal force, and the right rear wheel longitudinal force.

2. The integrated model prediction obstacle avoidance planning and control method according to claim 1, wherein Obtain the longitudinal speed of the vehicle , dimension information, safety threshold information; The dimension information includes the vehicle width of the host vehicle , the distance from the centroid of the host vehicle to the front of the vehicle , the width of the obstacle ; The safety threshold information includes the maximum lateral acceleration ; Based on the longitudinal speed of the vehicle and the dimension information, a vehicle obstacle avoidance behavior guidance trajectory formula is established through a fifth-degree polynomial; Obtain the obstacle avoidance behavior guidance trajectory based on the vehicle obstacle avoidance behavior guidance trajectory formula in combination with the safety threshold information ; Obtain the desired yaw angle of the vehicle through Formula 1 in combination with the vehicle obstacle avoidance trajectory information .

3. The integrated model predictive obstacle avoidance planning and control method according to claim 2, wherein The formula for the vehicle obstacle avoidance behavior guiding trajectory is: ； Among them, , , , , , ; ； ； , , , , , are the fitting coefficients of the obstacle avoidance trajectory curve; is the lateral displacement for obstacle avoidance; is the lateral reserved safety distance.

4. The integrated model prediction obstacle avoidance planning and control method according to claim 3, wherein The vehicle obstacle avoidance behavior guiding trajectory is: ； ； The longitudinal position of the vehicle's center of mass at the moment during the obstacle avoidance process, and the lateral position of the vehicle's center of mass at the moment during the obstacle avoidance process; The avoidance time, calculated as follows: ； Among them, ; .

5. The integrated model, obstacle avoidance planning and control method according to claim 4, characterized in that Trajectory guidance through obstacle avoidance behavior Combined with Formula 1 to obtain the desired yaw angle of the vehicle , and the Formula 1 is as follows: ； ； ； is the sampling time; is the lateral position change amount at adjacent times; is the longitudinal position change amount at adjacent times.

6. The integrated model prediction obstacle avoidance planning and control method according to claim 5, characterized in that, Maximum lateral acceleration including the maximum lateral acceleration for stable obstacle avoidance and the maximum lateral acceleration for emergency obstacle avoidance; The maximum value of the lateral acceleration for stable obstacle avoidance is ; The maximum value of the lateral acceleration for emergency obstacle avoidance is ; The said is the road adhesion coefficient; the said is the acceleration due to gravity; The autonomous driving vehicle autonomously switches to a smooth obstacle avoidance behavior or an emergency obstacle avoidance behavior according to the driver's personal habits and actual obstacle avoidance conditions; When the smooth obstacle avoidance behavior is selected, is the maximum value of the lateral acceleration for smooth obstacle avoidance, and the specific value is , which is substituted into the vehicle obstacle avoidance behavior guidance trajectory formula to obtain the smooth obstacle avoidance behavior guidance trajectory; When the emergency obstacle avoidance behavior is selected, is the maximum lateral acceleration for emergency obstacle avoidance, and the specific value is , which is substituted into the vehicle obstacle avoidance behavior guidance trajectory formula to obtain the emergency obstacle avoidance behavior guidance trajectory.

7. The integrated model predictive obstacle avoidance planning and control method according to claim 1, characterized in that The prediction model includes the following formula: ； ； Among them, ; ； ； ； is the front wheel steering angle, , , , are the longitudinal forces of the four wheels, , , , are the lateral forces of the four wheels, is the longitudinal resistance, is the additional yaw moment, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle of the vehicle, is the vehicle's track width, is the moment of inertia of the vehicle about the z-axis, is the vehicle mass, is the vehicle's longitudinal speed, is the vehicle's lateral speed, is the vehicle's yaw angle, is the vehicle's yaw angular velocity, is the longitudinal speed change rate, is the lateral speed change rate, is the longitudinal speed in the earth coordinate, is the lateral speed in the earth coordinate.

8. The integrated model prediction obstacle avoidance planning and control method according to claim 1, characterized in that The objective function is: ； is the sampling time; is the prediction horizon; is the control horizon; the first term of the objective function is the obstacle avoidance behavior guiding trajectory tracking accuracy term; in the first term of the objective function is the predicted value of the vehicle state information within the prediction horizon, and in the first term of the objective function is the expected value of the vehicle state information within the prediction horizon determined based on the obstacle avoidance behavior guiding trajectory; the second term is the control quantity term; the third term is the control quantity increment term; , , are three corresponding weight matrices; is the weight matrix corresponding to the obstacle avoidance behavior guidance trajectory tracking accuracy term, and the weight matrix includes the longitudinal speed tracking deviation weight, the vehicle lateral position tracking deviation weight, and the vehicle yaw angle tracking deviation weight; the longitudinal speed tracking deviation weight value is 2000 - 5000; the lateral position tracking deviation weight value is 5000 - 8000; the vehicle yaw angle tracking deviation weight value is 5000 - 8000; is the weight matrix corresponding to the control quantity item, and the weight matrix includes the front wheel steering angle weight and the wheel longitudinal force weight; the front wheel steering angle weight value is 50000 - 80000; the wheel longitudinal force weight value is 10000 - 20000; is the weight matrix corresponding to the control quantity increment term, and the weight matrix includes the weight of the front wheel steering angle increment and the weight of the wheel longitudinal force increment; the weight value of the front wheel steering angle increment is 50000 - 80000; the weight value of the wheel longitudinal force increment is 10000 - 20000.

9. The integrated model predictive obstacle avoidance planning and control method according to claim 1, characterized in that The control quantity constraint conditions include: ; The control quantity increment constraint conditions include: ; The control variables mainly include the front wheel steering angle and the longitudinal forces of the four wheels , , , ; , are the maximum and minimum values of the control variables respectively, determined by the physical limit boundaries of the actuator; , are the maximum and minimum values of the control variable increments respectively; The obstacle avoidance constraint requires that the vehicle is always within the drivable area, that is, the four vertices of the vehicle are within the safety boundary range to ensure the vehicle obstacle avoidance safety; the expression of the constraint condition is: ； ； ； ； ， ， ， ； , , , are the predicted lateral position values of the four vertices of the vehicle, is the lateral position of the vehicle's center of mass, is the longitudinal distance from the vehicle's center of mass to the front axle, is the longitudinal distance from the vehicle's center of mass to the rear axle, is the vehicle's track width, is the yaw angle of the vehicle; , , , , , , , are the obstacle avoidance safety boundary values corresponding to the four vertices of the vehicle, respectively.

10. The integrated model predictive obstacle avoidance planning and control method according to claim 1, characterized in that Solve the integrated non - linear optimization problem using sequential quadratic programming, and use the obtained sequence of optimal solutions as the initial solution for the next time step after moving one step forward , thus forming an iterative warm start; ； 。

Images