Multi-target collaborative optimization obstacle avoidance behavior trajectory decision method

Abstract

The invention discloses a multi-target collaborative optimization obstacle avoidance behavior trajectory decision-making method. The method comprises an obstacle avoidance behavior classification method, an obstacle avoidance trajectory fitting method and a multi-target comprehensive evaluation method. Through the obstacle avoidance behavior classification method, the obstacle avoidance behaviors are subdivided into a plurality of types of obstacle avoidance schemes according to a preset obstacle avoidance strategy; through the obstacle avoidance trajectory fitting method, performing fitting under a plurality of types of obstacle avoidance schemes to generate an obstacle avoidance behavior trajectory cluster; through a multi-target comprehensive evaluation method, a multi-target evaluation index is constructed based on vehicle actual state information and safety threshold information, and the closeness degree of each obstacle avoidance behavior track to an ideal solution is calculated based on the multi-target evaluation index. Selecting the obstacle avoidance behavior track with the highest proximity value as a decision result of the multi-target collaborative optimization obstacle avoidance behavior track; the multi-target comprehensive evaluation method comprises a safety evaluation index, a comfort evaluation index and an efficiency evaluation index; the optimal obstacle avoidance behavior track is obtained through decision making, and the obstacle avoidance safety, comfort and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle obstacle avoidance decision-making, and particularly to a field of an obstacle avoidance behavior trajectory decision-making method with multi-objective collaborative optimization. Background Art

[0002] In current production and life, assisted driving and autonomous driving have gradually integrated into life; the decision-making of obstacle avoidance behavior during driving is particularly important.

[0003] The goal of obstacle avoidance behavior planning is to make full use of perception information, timely select the optimal driving behavior, and provide a reference for subsequent obstacle avoidance planning and control. The current obstacle avoidance behavior decision-making methods include two categories: rule-based and learning-based. Their output results are mainly discrete driving behavior action instructions. Although the complex obstacle avoidance planning problem is positioned to the planning sub-problem of specific obstacle avoidance behaviors and provides a reference for obstacle avoidance actions, it does not effectively constrain the feasible solution space range of subsequent planning and control optimization iterations, and continuous guidance cannot be obtained during the trajectory planning and control process. Therefore, it cannot ensure that the actual obstacle avoidance behavior is consistent with the behavior decision result and effectively improve the operation real-time performance of obstacle avoidance trajectory planning and control.

[0004] Therefore, how to invent an obstacle avoidance behavior trajectory decision-making method, comprehensively consider obstacle avoidance strategies such as braking, steering, and variable-speed steering, generate diverse obstacle avoidance behavior trajectories, comprehensively evaluate the candidate obstacle avoidance behavior trajectories around obstacle avoidance safety, comfort, and efficiency, and determine the obstacle avoidance behavior trajectory with multi-objective collaborative optimization to provide effective guidance for obstacle avoidance planning and control has become a difficult problem urgently to be solved in this field. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides an obstacle avoidance behavior trajectory decision-making method with multi-objective collaborative optimization, obtains the optimal obstacle avoidance behavior trajectory, provides effective guidance for obstacle avoidance control, and comprehensively improves obstacle avoidance safety, comfort, and efficiency.

[0006] The present invention provides an obstacle avoidance behavior trajectory decision-making method with multi-objective collaborative optimization, including an obstacle avoidance behavior classification method, an obstacle avoidance trajectory fitting method, and a multi-objective comprehensive evaluation method; Through the obstacle avoidance behavior classification method, the obstacle avoidance behavior is subdivided into several types of obstacle avoidance schemes according to a preset obstacle avoidance strategy; Through the obstacle avoidance trajectory fitting method, an obstacle avoidance behavior trajectory cluster is fitted and generated under several types of obstacle avoidance schemes; Through the multi-objective comprehensive evaluation method, a multi-objective evaluation index is constructed based on the actual state information of the vehicle and the safety threshold information, the proximity degree of each obstacle avoidance behavior trajectory to the ideal solution is calculated based on the multi-objective evaluation index, and the obstacle avoidance behavior trajectory with the highest proximity degree value is selected as the result of the obstacle avoidance behavior trajectory decision-making with multi-objective collaborative optimization; The obstacle avoidance behavior classification method includes a preset obstacle avoidance strategy and an obstacle avoidance plan classification method; The preset obstacle avoidance strategy includes a braking obstacle avoidance strategy, a steering obstacle avoidance strategy, and a variable speed steering obstacle avoidance strategy; The braking obstacle avoidance strategy includes a smooth braking obstacle avoidance plan, an emergency braking obstacle avoidance plan, and a full braking obstacle avoidance plan; The steering obstacle avoidance strategy includes a smooth steering obstacle avoidance plan and an emergency steering obstacle avoidance plan; The variable speed steering obstacle avoidance strategy includes an accelerate - then - steer obstacle avoidance plan and a decelerate - then - steer obstacle avoidance plan; The multi - objective evaluation index includes a safety evaluation index, a comfort evaluation index, and an efficiency evaluation index.

[0007] Compared with the prior art, the present invention has the following beneficial effects: By including an obstacle avoidance behavior classification method, an obstacle avoidance trajectory fitting method, and a multi - objective comprehensive evaluation method, it is realized that through the obstacle avoidance behavior classification method, the preset obstacle avoidance strategy is refined into several types of plans, and then through the obstacle avoidance trajectory fitting method, an obstacle avoidance behavior trajectory cluster is generated by fitting under several types of obstacle avoidance plans, so as to obtain several obstacle avoidance behavior trajectories with obstacle avoidance plan category identifiers; Through the multi - objective comprehensive evaluation method, a multi - objective evaluation index is constructed based on the vehicle's actual state information and safety threshold information, and the proximity of each obstacle avoidance behavior trajectory to the ideal solution is calculated. Taking obstacle avoidance safety, comfort, and efficiency as evaluation dimensions, evaluation indexes are constructed for different types of obstacle avoidance behavior trajectories. Calculate the multi - objective evaluation values of each obstacle avoidance behavior trajectory and calculate the proximity to the ideal solution; By selecting the obstacle avoidance behavior trajectory with the highest proximity value as the result of the multi - objective collaborative optimization obstacle avoidance behavior trajectory decision, the optimal decision of the obstacle avoidance behavior trajectory is realized; The preset obstacle avoidance strategy includes a braking obstacle avoidance strategy, a steering obstacle avoidance strategy, and a variable speed steering obstacle avoidance strategy; the multi - objective evaluation index includes a safety evaluation index, a comfort evaluation index, and an efficiency evaluation index; Based on the generated obstacle avoidance behavior trajectory cluster, evaluation indexes are constructed around the three dimensions of safety, comfort, and efficiency, combined with the vehicle's actual state information and safety threshold information, to quantitatively evaluate the obstacle avoidance behavior trajectories, calculate the proximity of each obstacle avoidance behavior trajectory to the ideal solution, and select the obstacle avoidance behavior trajectory of multi - objective collaborative optimization based on this.

[0008] Further, the obstacle avoidance plans include smooth braking obstacle avoidance, emergency braking obstacle avoidance, full braking obstacle avoidance, smooth steering obstacle avoidance, emergency steering obstacle avoidance, accelerate - then - steer obstacle avoidance, and decelerate - then - steer obstacle avoidance; The obstacle avoidance plan classification method includes: During the obstacle avoidance process, when , , and is a smooth braking obstacle avoidance scheme; When , , and is an emergency braking obstacle avoidance scheme; When , , and is a full braking obstacle avoidance scheme; When , is a smooth steering obstacle avoidance scheme; When , is an emergency steering obstacle avoidance scheme; When , is a deceleration - then - steering obstacle avoidance scheme; When , is an acceleration - then - steering obstacle avoidance scheme; Among them is the longitudinal acceleration, is the lateral acceleration; The obstacle avoidance trajectory fitting method is piece - wise third - order Hermite interpolation. By fitting several obstacle avoidance schemes, an obstacle avoidance behavior trajectory cluster is obtained. The obstacle avoidance behavior cluster includes several obstacle avoidance behavior trajectories with obstacle avoidance scheme category identifiers.

[0009] Using the beneficial effects of the previous step to refine the preset obstacle avoidance strategy. On the one hand, it generates diverse obstacle avoidance behavior trajectories and enhances the completeness of obstacle avoidance behavior. On the other hand, there are differences in the calculation methods of evaluation indicators corresponding to different obstacle avoidance behaviors during multi - objective comprehensive evaluation. Through obstacle avoidance behavior classification, category identifiers can be formed, which is convenient for quickly determining the matching evaluation indicators.

[0010] Furthermore, the multi - objective comprehensive evaluation method includes: Calculating the critical safety distance between the host vehicle and the obstacle directly in front at the initial moment of obstacle avoidance , the critical safety distance between the host vehicle and the obstacle in the front - left at the initial moment of obstacle avoidance , the critical safety distance between the host vehicle and the obstacle in the rear - left at the initial moment of obstacle avoidance , and the total time consumed during the obstacle avoidance process ; The critical safety distance between the host vehicle and the obstacle directly in front at the initial moment of obstacle avoidance includes: the critical safety distance directly in front for smooth braking obstacle avoidance is , the critical safety distance directly in front for emergency braking obstacle avoidance , and the critical safety distance directly in front for full braking obstacle avoidance , Critical safe distance in the straight-ahead direction for obstacle avoidance during steady turning , Critical safe distance in the straight-ahead direction for obstacle avoidance during emergency turning , Critical safe distance in the straight-ahead direction for obstacle avoidance by accelerating first and then turning , Critical safe distance in the straight-ahead direction for obstacle avoidance by decelerating first and then turning ; , Critical safe distance between the host vehicle and the obstacle in the left-ahead direction at the initial moment of obstacle avoidance includes: Critical safe distance for left-ahead obstacle avoidance during steady turning , Critical safe distance for left-ahead obstacle avoidance during emergency turning ; Critical safe distance for left-ahead obstacle avoidance by accelerating first and then turning , Critical safe distance for left-ahead obstacle avoidance by decelerating first and then turning ; , Critical safe distance between the host vehicle and the obstacle in the left-rear direction at the initial moment of obstacle avoidance includes: Critical safe distance for left-rear obstacle avoidance during steady turning , Critical safe distance for left-rear obstacle avoidance during emergency turning , Critical safe distance for left-rear obstacle avoidance by accelerating first and then turning , Critical safe distance for left-rear obstacle avoidance by decelerating first and then turning ; , Total time taken for the obstacle avoidance process includes: Total time taken for obstacle avoidance by steady braking , Total time taken for obstacle avoidance by emergency braking , Total time taken for obstacle avoidance by full braking , Total time taken for obstacle avoidance by steady turning , Total time taken for obstacle avoidance by emergency turning , Total time taken for obstacle avoidance by accelerating first and then turning , Total time taken for obstacle avoidance by decelerating first and then turning .

[0011] Further, the critical safe distance in the straight-ahead direction for obstacle avoidance by braking is calculated by Formula 1, and Formula 1 is: ; , Total time taken for obstacle avoidance by braking is calculated by Formula 2, and Formula 2 is: ; , , ; is the initial vehicle speed of the host vehicle; is the speed of the obstacle ahead; is the longitudinal braking deceleration; is the response time of the braking system, ; is the braking force growth time, ; is the time for continuous braking until the vehicle stops; is the total time for braking to avoid obstacles; is the minimum longitudinal safety distance reserved from the vehicle after it stops to the obstacle in front; is calculated through Formula 1, ; is calculated through Formula 2, ; is calculated through Formula 1, ; is calculated through Formula 2, ; is calculated through Formula 1, ; is calculated through Formula 2, ;

[0012] The beneficial effect of adopting the previous step is to obtain the critical safety distance for smooth braking and obstacle avoidance in front, the critical safety distance for emergency braking and obstacle avoidance in front, the critical safety distance for full braking and obstacle avoidance in front ; the total time for smooth braking and obstacle avoidance , the total time for emergency braking and obstacle avoidance , the total time for full braking and obstacle avoidance ;

[0013] Furthermore, the critical safety distance for steering and obstacle avoidance in front is calculated through Formula 3, and Formula 3 is: ; The critical collision moment and the total time for steering and obstacle avoidance are calculated through Formula 4, and Formula 4 is: ; Among them, , , ,{t}_{sf}=\sqrt[{3}] {\frac {10\sqrt {3}×{Y}_{w}} {3×{a}_{ymax}}} , , ; The position of the vehicle's center of mass at the critical collision moment; is the vehicle body heading angle at the critical collision; The longitudinal safety distance reserved for obstacle avoidance; is the distance from the vehicle's center of mass to the front end of the vehicle; is the lateral safety distance reserved for obstacle avoidance; is the width of the vehicle body; The lateral boundary position of the obstacle directly ahead; is the maximum lateral acceleration; is the initial vehicle speed; The critical safety distance of the obstacle on the left front for steering obstacle avoidance is calculated by Formula Five, and Formula Five is: ; The critical safety distance of the obstacle on the left rear for steering obstacle avoidance is calculated by Formula Six, and Formula Six is: ; is the full braking distance; is the initial vehicle speed; is the traveling speed of the obstacle on the left front; is the traveling speed of the obstacle on the left rear.

[0014] Furthermore, through Formula Three, , through Formula Four, through Formula Five, through Formula Six, when is calculated; through Formula Three, , through Formula Four, through Formula Five, through Formula Six, when .

[0015] The beneficial effect of adopting the previous step is that the critical safety distance directly ahead for smooth steering obstacle avoidance under the steering obstacle avoidance strategy is obtained through the above content , the critical safety distance directly ahead for emergency steering obstacle avoidance , the critical safety distance of the obstacle on the left front for smooth steering obstacle avoidance , the critical safety distance of the obstacle on the left front for emergency steering obstacle avoidance , the critical safety distance of the obstacle on the left rear for smooth steering obstacle avoidance , the critical safety distance of the obstacle on the left rear for emergency steering obstacle avoidance , the total time consumption for smooth steering obstacle avoidance , the total time consumption for emergency steering obstacle avoidance calculation method.

[0016] Furthermore, the critical safety distance in the direct front during variable-speed steering to avoid obstacles is calculated by Formula 7, and Formula 7 is: ; wherein, , ; is the expected driving distance during the variable-speed stage; is the critical safety distance during the steering stage; is the initial vehicle speed of the host vehicle; is the target vehicle speed after variable speed; is the vehicle body heading angle at critical collision; is the distance from the center of mass of the host vehicle to the front end of the vehicle; is the width of the host vehicle body; is the width of the obstacle in the direct front; is the critical collision time during the steering process; is the acceleration or braking response time of the distributed drive vehicle, ; The total time consumed for variable-speed steering to avoid obstacles is calculated by Formula 8, and Formula 8 is as follows: ; The critical safety distance of the obstacle in the left front during variable-speed steering to avoid obstacles is calculated by Formula 9, and Formula 9 is: ; is the minimum braking distance; is the driving speed of the obstacle in the left front; The critical safety distance of the obstacle in the left rear during variable-speed steering to avoid obstacles is calculated by Formula 10, and Formula 10 is: ; is the speed of the obstacle in the left rear.

[0017] Furthermore, through Formula 7, through Formula 8, through Formula 9, through Formula 10, when is calculated; through Formula 7, through Formula 8, through Formula 9, through Formula 10, when is calculated.

[0018] The beneficial effect of adopting the previous step is that the critical safety distance in front of the obstacle when steering while changing speed to avoid the obstacle is obtained through the above content, accelerating first and then steering to avoid the obstacle , the critical safety distance in front of the obstacle when decelerating first and then steering to avoid the obstacle , the critical safety distance of the obstacle in the left front when accelerating first and then steering to avoid the obstacle , the critical safety distance of the obstacle in the left front when decelerating first and then steering , the critical safety distance of the obstacle in the left rear when accelerating first and then steering to avoid the obstacle , the critical safety distance of the obstacle in the left rear when decelerating first and then steering to avoid the obstacle , the total time taken when accelerating first and then steering to avoid the obstacle , the total time taken when decelerating first and then steering to avoid the obstacle calculation method

[0019] Furthermore, construct safety evaluation indicators , comfort evaluation indicators , efficiency evaluation indicators ; Calculate the safety evaluation index values corresponding to each obstacle avoidance behavior trajectory in the obstacle avoidance behavior trajectory cluster based on multi-objective evaluation indicators , comfort evaluation index values and efficiency evaluation index values ; Construct an evaluation index matrix , , ; is the number of obstacle avoidance behavior trajectories included in the obstacle avoidance behavior trajectory cluster; Evaluation index dimension, including safety evaluation index value, comfort evaluation index value, efficiency evaluation index value, ; For the evaluation index matrix perform normalization processing to obtain the normalized evaluation index matrix ; ; is any item in the evaluation index matrix; is the maximum value of the evaluation value corresponding to the jth evaluation index; is the minimum value of the evaluation value corresponding to the jth evaluation index; Combine the weights corresponding to the safety evaluation index, comfort evaluation index, and efficiency evaluation index For the evaluation index matrix perform weighted processing to obtain the weighted evaluation index matrix ; ; Determine the positive ideal value corresponding to each evaluation index and the negative ideal value ; , ; The positive ideal value is the maximum weighted evaluation value corresponding to the j-th evaluation index; the negative ideal value is the minimum weighted evaluation value corresponding to the j-th evaluation index; Determine the relative distances of each obstacle avoidance behavior trajectory from the positive ideal value and the negative ideal value through the Euclidean distance, so as to calculate the proximity of each obstacle avoidance behavior trajectory to the ideal solution; , , ; Among them, is the relative distance of the -th obstacle avoidance behavior trajectory from the positive ideal value; is the relative distance of the -th obstacle avoidance behavior trajectory from the negative ideal value; is the proximity value of the -th obstacle avoidance behavior trajectory to the ideal solution; Select the obstacle avoidance behavior trajectory with the largest proximity value as the result of the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision.

[0020] The beneficial effect of adopting the previous step is to comprehensively and quantitatively evaluate the obstacle avoidance behavior trajectory in terms of obstacle avoidance safety, comfort, and efficiency, determine the obstacle avoidance behavior trajectory of multi-objective collaborative optimization, and achieve the optimal decision of the obstacle avoidance behavior trajectory.

[0021] Furthermore, construct the safety evaluation index value , the comfort evaluation index value , and the efficiency evaluation index value based on the actual vehicle state information and the safety threshold information; ; is the obstacle avoidance urgency; is the collision damage mitigation ability; Obstacle avoidance possibility is the shifting risk; ; The actual distance between the host vehicle and the obstacle in front; is the critical safety distance between the host vehicle and the obstacle directly in front at the initial moment of obstacle avoidance; When is true, , ; When , , ; , ; is the degree of collision damage mitigation, determined by the difference between the potential collision severity at the end of obstacle avoidance and the potential collision severity at the initial moment of obstacle avoidance; is the initial vehicle speed of the host vehicle; is the equivalent speed at the collision moment, determined by the vehicle speed and body attitude of the host vehicle when the collision occurs; is the speed of the potential collision object; is the mass of the host vehicle; is the mass of the potential collision object; When adopting the braking obstacle avoidance strategy, ; When adopting the steering obstacle avoidance strategy or the variable speed steering obstacle avoidance strategy, ; is the actual distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; is the actual distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; is the critical safety distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; The critical safety distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; When adopting the braking obstacle avoidance strategy and the steering obstacle avoidance strategy, ; When adopting the accelerating steering obstacle avoidance strategy, , , , , ; When adopting the decelerating steering obstacle avoidance strategy, , , , , ; is the target vehicle speed after variable speed; is the minimum road speed limit; is the maximum road speed limit; is the actual deceleration or acceleration; and are the execution boundary values of the distributed drive vehicle , ; ; is the maximum longitudinal acceleration during obstacle avoidance, It is the maximum lateral acceleration during obstacle avoidance; is the longitudinal acceleration comfort threshold. , when using deceleration ; is the lateral acceleration comfort threshold, ; , ; is the road reference speed; is the average vehicle speed during the obstacle avoidance process; when the steering obstacle avoidance strategy is adopted, ; When using the braking obstacle avoidance strategy or the speed change steering obstacle avoidance strategy, ; The actual longitudinal distance traveled during the obstacle avoidance process; The total time consumed for obstacle avoidance; when the braking obstacle avoidance strategy is adopted, ; When the speed-changing steering obstacle avoidance strategy is adopted, .

[0022] The beneficial effect of adopting the previous step is that the safety evaluation index evaluates the collision risk based on the urgency of obstacle avoidance and the potential collision severity, and for the steering obstacle avoidance strategy and the speed change steering strategy, the possibility of lane change obstacle avoidance is evaluated based on the interference degree of obstacles in adjacent lanes. At the same time, for the speed change steering obstacle avoidance strategy, the target vehicle speed after the speed change and the acceleration of the speed change process are constrained based on the requirements of traffic safety regulations and the actual execution boundary to construct a speed change risk index; the comfort evaluation index comprehensively considers the impact of the longitudinal and lateral speed change rates on driving comfort, and minimizes the sense of frustration during the obstacle avoidance process; the efficiency index requires the vehicle to complete the obstacle avoidance action as soon as possible to minimize the obstruction to the surrounding traffic; and finally determines the obstacle avoidance behavior trajectory of multi-objective collaborative optimization, and provides effective guidance for subsequent obstacle avoidance control to comprehensively improve the safety, comfort and efficiency of obstacle avoidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 The present invention is a flow chart of the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision method including the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision method of Example 1. DETAILED DESCRIPTION

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

[0025] Embodiment 1: According to this embodiment, a multi - objective collaborative optimization obstacle - avoidance behavior trajectory decision - making method is provided, including an obstacle - avoidance behavior classification method, an obstacle - avoidance trajectory fitting method, and a multi - objective comprehensive evaluation method; The obstacle - avoidance behavior classification method includes a preset obstacle - avoidance strategy and an obstacle - avoidance scheme classification method; The preset obstacle - avoidance strategies include a braking obstacle - avoidance strategy, a steering obstacle - avoidance strategy, and a variable - speed steering obstacle - avoidance strategy; The braking obstacle - avoidance strategy includes a smooth braking obstacle - avoidance scheme, an emergency braking obstacle - avoidance scheme, and a full - braking obstacle - avoidance scheme; The steering obstacle - avoidance strategy includes a smooth steering obstacle - avoidance scheme and an emergency steering obstacle - avoidance scheme; The variable - speed steering obstacle - avoidance strategy includes a scheme of accelerating first and then steering to avoid obstacles and a scheme of decelerating first and then steering to avoid obstacles; The obstacle - avoidance scheme classification method includes: During the obstacle - avoidance process, when , , and it is a smooth braking obstacle - avoidance scheme; When , , and it is an emergency braking obstacle - avoidance scheme; When , , and it is a full - braking obstacle - avoidance scheme; When , it is a smooth steering obstacle - avoidance scheme; When , it is an emergency steering obstacle - avoidance scheme; When , it is a scheme of decelerating first and then steering to avoid obstacles; When , it is a scheme of accelerating first and then steering to avoid obstacles; where is the longitudinal acceleration, is the lateral acceleration.

[0026] Through the obstacle - avoidance trajectory fitting method, an obstacle - avoidance behavior trajectory cluster is fitted under several types of obstacle - avoidance schemes; the obstacle - avoidance trajectory fitting method adopts piece - wise third - order Hermite interpolation to fit several obstacle - avoidance behaviors to obtain an obstacle - avoidance behavior trajectory cluster, and the obstacle - avoidance behavior trajectory cluster includes several obstacle - avoidance behavior trajectories with obstacle - avoidance scheme category identifiers.

[0027] A multi-objective comprehensive evaluation method is used to construct a multi-objective evaluation index based on the actual vehicle status information and safety threshold information. The degree of proximity of each obstacle avoidance behavior trajectory to the ideal solution is calculated based on the multi-objective evaluation index, and the obstacle avoidance behavior trajectory with the highest degree of proximity is selected as the result of multi-objective collaborative optimization of obstacle avoidance behavior trajectory decision. A multi-objective comprehensive evaluation method is used to construct a multi-objective evaluation index based on the actual vehicle status information and safety threshold information. The degree of proximity of each obstacle avoidance behavior trajectory to the ideal solution is calculated based on the multi-objective evaluation index, and the obstacle avoidance behavior trajectory with the highest degree of proximity is selected as the result of multi-objective collaborative optimization of obstacle avoidance behavior trajectory decision. The multi-objective evaluation index includes a safety evaluation index, a comfort evaluation index and an efficiency evaluation index.

[0028] The multi-objective comprehensive evaluation method includes: Calculate the critical safety distance between the vehicle and the obstacle in front at the initial moment of obstacle avoidance , The critical safety distance between the main vehicle and the left front obstacle at the initial moment of obstacle avoidance , The critical safety distance between the main vehicle and the left rear obstacle at the initial moment of obstacle avoidance , the total time of obstacle avoidance process ; The critical safety distance between the vehicle and the obstacle in front at the initial moment of obstacle avoidance Includes: Smooth braking to avoid obstacles, the critical safety distance in front is , Emergency braking obstacle avoidance critical safety distance in front , Full braking obstacle avoidance critical safety distance in front , Steady steering to avoid obstacles and maintain critical safety distance in front , Emergency turn to avoid obstacles and critical safety distance in front , first accelerate and then turn to avoid obstacles directly in front of the critical safety distance , first slow down and then turn to avoid the critical safety distance in front of the obstacle ; The critical safety distance between the main vehicle and the left front obstacle at the initial moment of obstacle avoidance Includes: Smooth steering, obstacle avoidance, critical safety distance to left front obstacle , Emergency turn to avoid obstacles on the left front, critical safety distance ; Accelerate first and then turn to avoid the obstacle in front of the left critical safety distance , first slow down and then turn to avoid the obstacle on the left front. Critical safety distance ; The critical safety distance between the main vehicle and the left rear obstacle at the initial moment of obstacle avoidance Includes: Smooth steering, obstacle avoidance, left rear obstacle, critical safety distance , Emergency steering to avoid obstacles on the left rear, critical safety distance 、Critical safety distance for avoiding obstacles on the left rear by accelerating first and then steering 、Critical safety distance for avoiding obstacles on the left rear by decelerating first and then steering ; Total time taken for the obstacle avoidance process Includes: Total time taken for obstacle avoidance by smooth braking 、Total time taken for obstacle avoidance by emergency braking 、Total time taken for obstacle avoidance by full braking 、Total time taken for obstacle avoidance by smooth steering 、Total time taken for obstacle avoidance by emergency steering 、Total time taken for obstacle avoidance by accelerating first and then steering 、Total time taken for obstacle avoidance by decelerating first and then steering 。

[0029] Critical safety distance for braking obstacle avoidance directly ahead Calculated by Formula 1, and Formula 1 is: ; Total time taken for braking obstacle avoidance Calculated by Formula 2, and Formula 2 is: ; , , ; is the initial vehicle speed; is the speed of the obstacle ahead; is the longitudinal braking deceleration; is the braking system response time, ; is the braking force growth time, ; is the time for continuous braking until stopping; is the total time taken for braking obstacle avoidance; is the minimum longitudinal safety distance reserved from the vehicle ahead after the vehicle stops; Through Formula 1, Calculated; Through Formula 2, Calculated; Through Formula 1, Calculated, Through Formula 2, Calculated; Through Formula 1, Calculated, Through Formula 2, Calculated

[0030] Critical safety distance in the direct front of steering obstacle avoidance Calculated by Formula 3, and Formula 3 is: ; Critical collision moment and total time of steering obstacle avoidance Calculated by Formula 4, and Formula 4 is: ; wherein, , , , {t}_{sf}=\sqrt[{3}] {\frac {10\sqrt {3}×{Y}_{w}} {3×{a}_{ymax}}} , , ; Position of the center of mass of the host vehicle at the critical collision moment; is the vehicle body heading angle at the critical collision; Longitudinal reserved safety distance for obstacle avoidance; is the distance from the center of mass of the host vehicle to the front end of the vehicle; is the lateral reserved safety distance for obstacle avoidance; is the width of the host vehicle body; Lateral boundary position of the obstacle in the direct front; is the maximum lateral acceleration; is the initial vehicle speed of the host vehicle; Critical safety distance of the obstacle in the left front of steering obstacle avoidance Calculated by Formula 5, and Formula 5 is: ; Critical safety distance of the obstacle in the left rear of steering obstacle avoidance Calculated by Formula 6, and Formula 6 is: ; is the full braking distance; is the initial vehicle speed of the host vehicle; is the driving speed of the obstacle in the left front; is the driving speed of the obstacle in the left rear.

[0031] Calculated by Formula 3, , Calculated by Formula 4, Calculated by Formula 5, Calculated by Formula 6 when ; Calculated by Formula 3, , Through Formula Four, Through Formula Five, Through Formula Six, when .

[0032] Critical safety distance in the dead ahead during variable-speed steering obstacle avoidance Is calculated through Formula Seven, and Formula Seven is: ; Wherein, , ; Is the expected driving distance during the variable-speed stage; Is the critical safety distance during the steering stage; Is the initial vehicle speed of the host vehicle; Is the target vehicle speed after variable speed; Is the vehicle body heading angle at critical collision; Is the distance from the center of mass of the host vehicle to the front end of the vehicle; Is the width of the host vehicle body; Is the width of the dead-ahead obstacle; Is the critical collision time during the steering process; Is the acceleration or braking response time of the distributed drive vehicle, ; Total time-consuming of variable-speed steering obstacle avoidance Is calculated through Formula Eight, and Formula Eight is as follows: ; Critical safety distance of the left-ahead obstacle during variable-speed steering obstacle avoidance Is calculated through Formula Nine, and Formula Nine is: ; Is the minimum braking distance; Is the driving speed of the left-ahead obstacle; Critical safety distance of the left-rear obstacle during variable-speed steering obstacle avoidance Is calculated through Formula Ten, and Formula Ten is: ; Is the speed of the left-rear obstacle.

[0033] Through Formula Seven, Through Formula Eight, Through Formula Nine, Through Formula Ten, when Is calculated; Through Formula Seven, Through Formula Eight, Through Equation Nine, Through Equation Ten, when it is calculated.

[0034] Construct safety evaluation indicators , comfort evaluation indicators , and efficiency evaluation indicators ; Based on the multi-objective evaluation indicators, calculate the safety evaluation indicator values, comfort evaluation indicator values , and efficiency evaluation indicator values corresponding to each obstacle avoidance behavior trajectory in the obstacle avoidance behavior trajectory cluster; ; Based on the actual vehicle state information and safety threshold information, construct safety evaluation indicator values , comfort evaluation indicator values , and efficiency evaluation indicator values ; ; is the obstacle avoidance urgency; is the collision injury mitigation ability; Obstacle avoidance possibility; is the shifting risk; ; The actual distance between the host vehicle and the obstacle ahead; is the critical safety distance between the host vehicle and the obstacle directly ahead at the initial moment of obstacle avoidance; When , , ; when , , ; , ; is the degree of collision injury mitigation, determined by the difference between the potential collision severity at the end of obstacle avoidance and the potential collision severity at the initial moment of obstacle avoidance; is the initial vehicle speed of the host vehicle; is the equivalent speed at the collision moment, determined by the vehicle speed and body attitude of the host vehicle at the time of collision; is the speed of the potential collision object; is the mass of the host vehicle; is the mass of the potential collision object; When the braking obstacle avoidance strategy is adopted, ; When adopting the steering obstacle avoidance strategy or the variable-speed steering obstacle avoidance strategy, ; is the actual distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; is the actual distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; is the critical safety distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; The critical safety distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; When adopting the braking obstacle avoidance strategy or the steering obstacle avoidance strategy, ; When adopting the accelerating steering obstacle avoidance strategy, , , , , ; When adopting the decelerating steering obstacle avoidance strategy, , , , , ; is the target vehicle speed after variable speed; is the minimum speed limit of the road; is the maximum speed limit of the road; is the actual deceleration or acceleration; and are the execution boundary values of the distributed drive vehicle , ; ; is the maximum value of the longitudinal acceleration during the obstacle avoidance process, is the maximum value of the lateral acceleration during the obstacle avoidance process; is the comfort threshold of the longitudinal acceleration. When accelerating, , when decelerating, ; is the comfort threshold of the lateral acceleration, ; , ; is the reference vehicle speed of the road; is the average vehicle speed during the obstacle avoidance process. When adopting the steering obstacle avoidance strategy, ; When adopting the braking obstacle avoidance strategy or the variable-speed steering obstacle avoidance strategy, ; is the actual longitudinal driving distance during the obstacle avoidance process; is the total time for obstacle avoidance; when using the braking obstacle avoidance strategy, ; when using the variable speed and steering obstacle avoidance strategy, .

[0035] Construct an evaluation index matrix , , ; is the number of obstacle avoidance behavior trajectories included in the obstacle avoidance behavior trajectory cluster; The evaluation index dimension includes the safety evaluation index value, the comfort evaluation index value, and the efficiency evaluation index value, ; For the evaluation index matrix perform normalization processing to obtain the normalized evaluation index matrix ; ; is any item in the evaluation index matrix; is the maximum value of the evaluation value corresponding to the j-th evaluation index; is the minimum value of the evaluation value corresponding to the j-th evaluation index; Combined with the weights corresponding to the safety evaluation index, the comfort evaluation index, and the efficiency evaluation index For the evaluation index matrix perform weighted processing to obtain the weighted evaluation index matrix ; ; Determine the positive ideal value and the negative ideal value ; , ; The positive ideal value is the maximum weighted evaluation value corresponding to the j-th evaluation index; the negative ideal value is the minimum weighted evaluation value corresponding to the j-th evaluation index; Determine the relative distances of each obstacle avoidance behavior trajectory from the positive ideal value and the negative ideal value through the Euclidean distance, so as to calculate the degree of closeness of each obstacle avoidance behavior trajectory to the ideal solution; , , ; Among them, is the relative distance of the -th obstacle avoidance behavior trajectory from the positive ideal value; is the relative distance of the th obstacle avoidance behavior trajectory from the negative ideal value; is the proximity value of the th obstacle avoidance behavior trajectory to the ideal solution; Select the obstacle avoidance behavior trajectory with the largest proximity value as the result of the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision.

[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. A method for obstacle avoidance behavior trajectory decision-making with multi-objective collaborative optimization, characterized in that it includes an obstacle avoidance behavior classification method, an obstacle avoidance trajectory fitting method, and a multi-objective comprehensive evaluation method; through the obstacle avoidance behavior classification method, the obstacle avoidance behavior is subdivided into several obstacle avoidance schemes according to a preset obstacle avoidance strategy; through the obstacle avoidance trajectory fitting method, an obstacle avoidance behavior trajectory cluster is generated by fitting under several obstacle avoidance schemes; through the multi-objective comprehensive evaluation method, a multi-objective evaluation index is constructed based on the actual vehicle state information and safety threshold information, the proximity of each obstacle avoidance behavior trajectory to the ideal solution is calculated based on the multi-objective evaluation index, and the obstacle avoidance behavior trajectory with the highest proximity value is selected as the result of the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision-making; the obstacle avoidance behavior classification method includes a preset obstacle avoidance strategy and an obstacle avoidance scheme classification method; the preset obstacle avoidance strategy includes a braking obstacle avoidance strategy, a steering obstacle avoidance strategy, and a variable-speed steering obstacle avoidance strategy; the braking obstacle avoidance strategy includes a smooth braking obstacle avoidance scheme, an emergency braking obstacle avoidance scheme, and a full braking obstacle avoidance scheme; the steering obstacle avoidance strategy includes a smooth steering obstacle avoidance scheme and an emergency steering obstacle avoidance scheme; the variable-speed steering obstacle avoidance strategy includes an accelerating-then-turning obstacle avoidance scheme and a decelerating-then-turning obstacle avoidance scheme; the multi-objective evaluation index includes a safety evaluation index, a comfort evaluation index, and an efficiency evaluation index.

2. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 1, wherein The obstacle avoidance schemes include smooth braking obstacle avoidance, emergency braking obstacle avoidance, full braking obstacle avoidance, smooth steering obstacle avoidance, emergency steering obstacle avoidance, accelerating-then-turning obstacle avoidance, and decelerating-then-turning obstacle avoidance; The obstacle avoidance scheme classification method includes: During the obstacle avoidance process, when , , and it is a stable braking obstacle avoidance scheme; When , , and is an emergency braking obstacle avoidance scheme; When , , and is a full braking obstacle avoidance scheme; When , it is a smooth steering obstacle avoidance solution; When , is an emergency steering obstacle avoidance solution; When , is a deceleration-first and then steering obstacle avoidance solution; When , is the first accelerating and then turning obstacle avoidance scheme; wherein is the longitudinal acceleration, is the lateral acceleration.

3. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 2, characterized in that The multi-objective comprehensive evaluation method includes: Calculate the critical safety distance between the host vehicle and the obstacle directly in front at the initial moment of obstacle avoidance and the critical safety distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance and the critical safety distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance and the total time taken for the obstacle avoidance process ; The critical safety distance between the host vehicle and the obstacle directly in front at the initial moment of obstacle avoidance including: the critical safety distance for obstacle avoidance by smooth braking directly in front is , the critical safety distance for obstacle avoidance by emergency braking directly in front , the critical safety distance for obstacle avoidance by full braking directly in front , the critical safety distance for obstacle avoidance by smooth steering directly in front , the critical safety distance for obstacle avoidance by emergency steering directly in front , the critical safety distance for obstacle avoidance by accelerating first and then steering directly in front , the critical safety distance for obstacle avoidance by decelerating first and then steering directly in front ; The critical safety distance between the host vehicle and the obstacle in the upper left front at the initial moment of obstacle avoidance including: the critical safety distance for smoothly steering to avoid the obstacle in the upper left front , the critical safety distance for emergently steering to avoid the obstacle in the upper left front ; the critical safety distance for accelerating first and then steering to avoid the obstacle in the upper left front , the critical safety distance for decelerating first and then steering to avoid the obstacle in the upper left front ; Critical safety distance between the host vehicle and the obstacle at the left rear at the initial moment of obstacle avoidance Including: Critical safety distance for smoothly steering to avoid the obstacle at the left rear , Critical safety distance for emergently steering to avoid the obstacle at the left rear , Critical safety distance for accelerating first and then steering to avoid the obstacle at the left rear , Critical safety distance for decelerating first and then steering to avoid the obstacle at the left rear ; Total time taken for obstacle avoidance process Including: total time taken for smooth braking obstacle avoidance , total time taken for emergency braking obstacle avoidance , total time taken for full braking obstacle avoidance , total time taken for smooth steering obstacle avoidance , total time taken for emergency steering obstacle avoidance , total time taken for accelerating first then steering obstacle avoidance , total time taken for decelerating first then steering obstacle avoidance .

4. The method for obstacle avoidance behavior trajectory decision-making with multi-objective collaborative optimization according to claim 3, wherein Critical safety distance for braking and obstacle avoidance directly ahead Obtained by calculating according to Formula 1, and Formula 1 is as follows: ； Total time for braking and obstacle avoidance Calculated by Formula 2, and Formula 2 is as follows: ； ， ， ； is the initial vehicle speed; is the speed of the obstacle ahead; is the longitudinal braking deceleration; is the braking system response time, ; is the braking force growth time, ; is the time for continuous braking until stopping; is the total time consumed for braking to avoid obstacles; is the minimum longitudinal safety distance reserved from the front obstacle after the vehicle stops; Obtained by Formula 1, Calculated; Obtained by Formula 2, Calculated; Through Formula 1, it is calculated that Through Formula 2, it is calculated; Through Formula 1, it is calculated that Through Formula 2, it is calculated.

5. The multi-objective collaborative optimization obstacle avoidance behavior trajectory decision-making method according to claim 3, characterized in that Critical safety distance in the direct front for steering to avoid obstacles It is calculated by Formula 3, and Formula 3 is as follows: ; Critical collision moment and the total time of steering and obstacle avoidance Calculated by Formula 4, and Formula 4 is: ； Among them, , , , , , ; The position of the center of mass of the host vehicle at the critical collision moment; is the vehicle body heading angle at the critical collision; The longitudinal safety distance reserved for obstacle avoidance; is the distance from the center of mass of the host vehicle to the front end of the vehicle; is the lateral safety distance reserved for obstacle avoidance; is the width of the host vehicle body; The lateral boundary position of the obstacle directly in front; is the maximum lateral acceleration; is the initial vehicle speed of the host vehicle; Critical safety distance of the left front obstacle for steering and obstacle avoidance Obtained by calculating according to Formula Five, and Formula Five is as follows: ; Critical safety distance of left rear obstacle for steering and obstacle avoidance Calculated by Formula 6, and Formula 6 is as follows: ; is the full braking distance; is the initial vehicle speed; is the driving speed of the obstacle in the left front; is the driving speed of the obstacle in the left rear.

6. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 5, characterized in that Through formula three, , Through formula four, Through formula five, Through formula six, when is calculated as; Through formula three, , Through formula four, Through formula five, Through formula six, when .

7. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 3, characterized in that, Critical safety distance in the direct front for variable-speed steering to avoid obstacles Obtained by calculating with Formula Seven, where Formula Seven is: ； Among them, , ; is the expected driving distance during the speed change phase; is the critical safety distance during the steering phase; is the initial vehicle speed of the host vehicle; is the target vehicle speed after the speed change; is the body heading angle at critical collision; is the distance from the center of mass of the host vehicle to the front end of the vehicle; is the body width of the host vehicle; is the width of the obstacle directly ahead; is the critical collision time during the steering process; is the acceleration or braking response time of the distributed drive vehicle, ; Total time taken for variable-speed steering to avoid obstacles Calculated using Equation VIII, which is as follows: ； Critical safety distance of the left front obstacle for variable-speed steering to avoid obstacles Obtained by calculating with Formula 9, and Formula 9 is as follows: ； is the minimum braking distance; is the driving speed of the obstacle in the front left; Critical safety distance for avoiding obstacles on the left rear during variable-speed steering Obtained by calculating using Equation Ten, and Equation Ten is as follows: ； is the speed of the obstacle at the left rear.

8. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 7, characterized in that Through formula seven, Through formula eight, Through formula nine, Through formula ten, when Calculated; Through Formula VII, Through Formula VIII, Through Formula IX, Through Formula X, when is calculated.

9. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 2, wherein, Construct safety evaluation indicators , comfort evaluation indicators , efficiency evaluation indicators ; Calculate the safety evaluation index value, comfort evaluation index value, and efficiency evaluation index value corresponding to each obstacle avoidance behavior trajectory in the obstacle avoidance behavior trajectory cluster based on multi-objective evaluation indicators , comfort evaluation index value and efficiency evaluation index value ; Construct an evaluation index matrix , , ; The number of obstacle avoidance behavior trajectories included in the obstacle avoidance behavior trajectory cluster; Evaluation index dimensions, including safety evaluation index values, comfort evaluation index values, and efficiency evaluation index values, ; For the evaluation index matrix perform normalization processing to obtain the normalized evaluation index matrix ; ； is any item in the evaluation index matrix; is the maximum value of the evaluation value corresponding to the j-th evaluation index; is the minimum value of the evaluation value corresponding to the j-th evaluation index; Combine the weights corresponding to the safety evaluation index, comfort evaluation index, and efficiency evaluation index For the evaluation index matrix Perform weighted processing to obtain a weighted evaluation index matrix ; ； Determine the positive ideal value corresponding to each evaluation index and the negative ideal value ; ， ； Positive ideal value is the maximum weighted evaluation value corresponding to the j-th evaluation index; Negative ideal value is the minimum weighted evaluation value corresponding to the j-th evaluation index; Determine the relative distances of each obstacle avoidance behavior trajectory from the positive ideal value and the negative ideal value through the Euclidean distance, so as to calculate the proximity of each obstacle avoidance behavior trajectory to the ideal solution; ， ， ； Among them, is the relative distance of the th obstacle avoidance behavior trajectory from the positive ideal value; is the relative distance of the th obstacle avoidance behavior trajectory from the negative ideal value; is the proximity value of the th obstacle avoidance behavior trajectory to the ideal solution; Select the obstacle avoidance behavior trajectory with the largest proximity value as the result of the multi-objective collaborative optimization obstacle avoidance behavior trajectory decision-making.

10. The obstacle avoidance behavior trajectory decision-making method for multi-objective collaborative optimization according to claim 9, characterized in that, Construct safety evaluation index values based on vehicle actual state information and safety threshold information , comfort evaluation index values , efficiency evaluation index values ; ； is the obstacle avoidance urgency; is the collision damage mitigation ability; is the obstacle avoidance possibility; is the speed change risk; ； The actual distance between the host vehicle and the obstacle ahead is the critical safety distance between the host vehicle and the obstacle directly ahead at the initial moment of obstacle avoidance When then , ; When then , ; ， ； is the degree of collision damage mitigation, determined by the difference between the potential collision severity at the end of obstacle avoidance and the potential collision severity at the initial moment of obstacle avoidance; is the initial vehicle speed; is the equivalent speed at the collision moment, determined by the vehicle speed and body attitude at the time of collision; is the speed of the potential collision object; is the vehicle mass; is the mass of the potential collision object;​ When adopting the braking obstacle avoidance strategy, ; When adopting the steering obstacle avoidance strategy or the variable-speed steering obstacle avoidance strategy, ; The actual distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; The actual distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; The critical safety distance between the host vehicle and the obstacle in the front left at the initial moment of obstacle avoidance; The critical safety distance between the host vehicle and the obstacle in the rear left at the initial moment of obstacle avoidance; When adopting the braking obstacle avoidance strategy and the steering obstacle avoidance strategy, ; When adopting the accelerated steering obstacle avoidance strategy, , , , , ; When adopting the deceleration steering obstacle avoidance strategy, , , , , ; is the target vehicle speed after speed change; is the minimum road speed limit; is the maximum road speed limit; is the actual deceleration or acceleration; and are the boundary values for the distributed drive vehicle to execute , ; ； is the maximum longitudinal acceleration during the obstacle avoidance process, is the maximum lateral acceleration during the obstacle avoidance process; is the longitudinal acceleration comfort threshold when acceleration is adopted and is when deceleration is adopted ; is the lateral acceleration comfort threshold, ; ， ； is the reference vehicle speed for the road; is the average vehicle speed during the obstacle avoidance process; when adopting the steering obstacle avoidance strategy, ; When adopting the braking obstacle avoidance strategy or the variable speed steering obstacle avoidance strategy, ; is the actual longitudinal driving distance during the obstacle avoidance process; is the total time consumed for obstacle avoidance; when adopting the braking obstacle avoidance strategy, ; When adopting the variable-speed steering obstacle avoidance strategy, .

Images