A trajectory planning method in a lane changing process of an autonomous vehicle, a storage medium and an equipment

Abstract

A trajectory planning method in a lane changing process of an autonomous vehicle, a storage medium and equipment, the present application relates to trajectory planning method in a lane changing process of an autonomous vehicle, a storage medium and equipment.The purpose of the present application is to solve the problem of low accuracy of lane changing feasibility discrimination and poor universality of lane changing necessity discrimination rule in the existing lane changing process of autonomous vehicle.The process is: the longitudinal displacement of the autonomous vehicle is determined;the game rule of the game participant is established, the optimal decision of the autonomous vehicle meeting the constraint condition is calculated according to the reverse induction method, the lateral displacement of the autonomous vehicle at the next time is determined;the trajectory sequence point of the autonomous vehicle at the next time is determined;the total time length of lane changing is set, the total benefit of the autonomous vehicle is obtained, the lane corresponding to the maximum total benefit is selected as the target lane, and the optimal decision of the autonomous vehicle is executed every time to realize the final lane changing trajectory of the autonomous vehicle.The present application is used in the field of trajectory planning.

Description

Technical Field

[0001] The present invention relates to a trajectory planning method, storage medium, and device for an autonomous driving vehicle during a lane change process. Background Art

[0002] The global autonomous driving technology sector is currently experiencing positive development. Relying on advanced sensor, communication, intelligent decision-making, and control technologies, autonomous vehicles are gradually upgrading their active safety and intelligent decision-making capabilities. Trajectory planning, in particular, has enabled autonomous vehicles to make the leap from state perception to intelligent control.

[0003] Trajectory planning focuses on the real-time transformation of vehicle states, while also considering vehicle dynamics, obstacle constraints, and risk assessment of driving operations. Currently, domestic and international scholars have conducted extensive research on trajectory planning for lane changes for autonomous vehicles, but most of these methods are developed for scenarios involving obstacle avoidance or forced lane changes, such as on- and off-ramps. Specific technical details involve utility maximization theory, risk field theory, quadratic programming, and machine learning. For rule-based models, designing transition conditions for driving scenarios becomes increasingly difficult when traffic scenarios become complex, making the model more complex. Machine learning-based methods, on the other hand, can only guarantee statistical optimality, but lack inherent interpretability and generalizability.

[0004] In reality, when autonomous vehicles travel on structured highways, in addition to forced lane changes, they will also change lanes to improve the driving environment. During this decision-making process, they compare the benefits before and after changing lanes and execute the decision that maximizes their own benefits. To this end, we propose a simple and highly generalizable trajectory planning method for autonomous vehicles during lane changes by improving the Stackelberg game model and supplementing it with the IDM model. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of low accuracy in judging the feasibility of lane changing and poor universality of the rules for judging the necessity of lane changing during the existing autonomous driving vehicle lane changing process, and to propose a trajectory planning method, storage medium and device for the lane changing process of an autonomous driving vehicle.

[0006] A trajectory planning method for an autonomous vehicle during lane change is as follows:

[0007] Step 1: Obtain the self-driving vehicle SV's own motion information and the autonomous driving vehicle SV's geometric dimensions;

[0008] Acquire geometric dimensions of vehicles Z surrounding the autonomous driving vehicle and motion state information of vehicles Z surrounding the autonomous driving vehicle;

[0009] Step 2: Based on the autonomous driving vehicle's own motion information from step 1 and the geometric dimensions and motion state information of vehicles surrounding the autonomous driving vehicle, a longitudinal displacement decision is made;

[0010] Step 3: Establish the game rules for the game participants. The autonomous vehicle is the leader and needs to make the first decision. The vehicles behind it in the lanes on both sides are followers and respond based on the leader's decision.

[0011] Based on the Follower's response to the Leader's decision, the optimal decision of the autonomous vehicle that satisfies the constraints is calculated using the backward induction method. The lateral displacement of the autonomous vehicle at the next moment is determined based on the optimal decision.

[0012] Step 4: Determine the trajectory sequence points of the autonomous driving vehicle at the next moment based on the lateral displacement and longitudinal displacement;

[0013] Step 5: Set the total lane change time to L seconds, repeat steps 1 to 4 L / Δt times, and obtain the corresponding total revenue of the autonomous driving vehicle U total , select the total revenue U of the autonomous driving vehicle total The lane corresponding to the maximum value is the target lane. Each time the autonomous driving vehicle executes the optimal decision to achieve the final lane changing trajectory of the autonomous driving vehicle.

[0014] A computer storage medium stores at least one instruction, which is loaded and executed by a processor to implement a trajectory planning method for an autonomous vehicle during lane changing.

[0015] A trajectory planning device for an autonomous vehicle during lane changing, the device comprising a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement a trajectory planning method for an autonomous vehicle during lane changing.

[0016] The beneficial effects of the present invention are:

[0017] The present invention proposes a trajectory planning method for an autonomous vehicle during a lane change process. The present invention improves the Stackelberg game model in economics, uses the advanced sensor technology of the autonomous vehicle to perceive the status information of surrounding vehicles, and takes into account the potential reactions of surrounding vehicles to the autonomous vehicle's decision, so that the autonomous vehicle can adjust its driving strategy in a timely manner, determine the optimal time to change lanes, and output a lateral displacement that takes both safety and efficiency into account. At the same time, the IDM model is used to plan the longitudinal displacement for the autonomous vehicle, ultimately realizing trajectory planning for the autonomous vehicle during lane change, improving the accuracy of lane change feasibility judgment during the existing autonomous vehicle lane change process, and enhancing the universality of the lane change necessity judgment rules during the existing autonomous vehicle lane change process, opening up a new path for future trajectory planning for autonomous vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a trajectory planning method for an autonomous driving vehicle during lane changing according to the present invention.

[0019] Figure 2 This is a schematic diagram of a vehicle trajectory during a lane changing process of an autonomous driving vehicle according to the present invention. DETAILED DESCRIPTION

[0020] Specific embodiment 1: This embodiment is a trajectory planning method for an autonomous driving vehicle during lane change. The specific process is as follows:

[0021] Step 1: Obtain the self-driving vehicle SV's own motion information and the autonomous driving vehicle SV's geometric dimensions;

[0022] Obtaining geometric dimensions of vehicles Z surrounding the autonomous driving vehicle and motion state information of vehicles Z surrounding the autonomous driving vehicle;

[0023] Step 2: Based on the autonomous driving vehicle's own motion information from step 1 and the geometric dimensions and motion state information of vehicles surrounding the autonomous driving vehicle, a longitudinal displacement decision is made;

[0024] Step 3: Establish the game rules for the game participants. The autonomous vehicle is the leader and needs to make the first decision. The vehicles behind it in the lanes on both sides are followers, and the followers respond based on the leader's decision (steps 21 to 24).

[0025] Based on the Follower's response to the Leader's decision, the optimal decision of the autonomous vehicle that satisfies the constraints is calculated using the backward induction method. The lateral displacement of the autonomous vehicle at the next moment is determined based on the optimal decision.

[0026] Step four: determining the trajectory sequence point of the autonomous vehicle at the next time according to the lateral displacement and the longitudinal displacement;

[0027] Step five: setting the total lane changing time as L seconds, repeating steps one to four L / Δt times, and obtaining the total benefit U of the autonomous vehicle total corresponding to the maximum value of the total benefit U of the autonomous vehicle total The lane corresponding to the maximum value is the target lane, and the autonomous vehicle implements the optimal decision each time to realize the final lane changing trajectory of the autonomous vehicle.

[0028] Specific implementation method two: different from the specific implementation method one is that in the step one, the motion information of the autonomous vehicle SV itself and the geometric size of the autonomous vehicle SV are obtained.

[0029] The geometric size of the surrounding vehicle Z of the autonomous vehicle and the motion state information of the surrounding vehicle Z of the autonomous vehicle are obtained.

[0030] The specific process is as follows:

[0031] The autonomous vehicle obtains the motion information of the autonomous vehicle SV itself and the geometric size of the autonomous vehicle SV through a vehicle-mounted computing unit.

[0032] The motion information of the autonomous vehicle SV itself includes speed v SV and acceleration acc SV .

[0033] The geometric size of the autonomous vehicle SV includes vehicle length l SV and vehicle width w SV .

[0034] The autonomous vehicle obtains the geometric size of the surrounding vehicle Z of the autonomous vehicle and the motion state information of the surrounding vehicle Z of the autonomous vehicle through a vehicle-mounted detection unit.

[0035] The geometric size of the surrounding vehicle Z of the autonomous vehicle includes vehicle length l Z and vehicle width w Z .

[0036] The motion state information of the surrounding vehicle Z of the autonomous vehicle includes speed v Z , acceleration acc Z , longitudinal relative distance Δy Z between the autonomous vehicle and the surrounding vehicle Z (calculated at the midpoint of the front bumper of the vehicle), distance d SV,1 from the autonomous vehicle to the left boundary of the lane in the driving direction, and distance d SV,2 from the autonomous vehicle to the right boundary of the lane in the driving direction.

[0037] Wherein, Z e {LV1, LV2, LV3, FV1, FV2}, LV1 represents the closest front vehicle of the automatic driving vehicle in the left lane of the automatic driving vehicle; LV2 represents the closest front vehicle of the automatic driving vehicle in the current lane of the automatic driving vehicle; LV3 represents the closest front vehicle of the automatic driving vehicle in the right lane of the automatic driving vehicle; FV1 represents the closest rear vehicle of the automatic driving vehicle in the left lane of the automatic driving vehicle; FV2 represents the closest rear vehicle of the automatic driving vehicle in the right lane of the automatic driving vehicle.

[0038] The above information will be input for the subsequent behavior decision of the automatic driving vehicle.

[0039] The other steps and parameters are the same as those in the first embodiment.

[0040] The third embodiment is different from the first or second embodiment in that the step two is based on the motion information of the automatic driving vehicle itself and the geometric size and motion state information of the surrounding vehicles of the automatic driving vehicle to make a longitudinal displacement decision of the automatic driving vehicle.

[0041] The specific process is as follows:

[0042] The IDM (Intelligent Driver Model) model is used to calculate the longitudinal displacement Δy(k) of the automatic driving vehicle.

[0043]

[0044] v SV (k) = v SV (k-1) + acc SV (k-1) x Δt

[0045]

[0046] Wherein, acc SV (k) represents the acceleration of the automatic driving vehicle at time k, and the initial value is 0; a IDM is the maximum acceleration of the automatic driving vehicle; v SV (k) is the speed of the automatic driving vehicle at time k, and the initial value is 0; is the speed of LV2 at time k; v des is the desired speed of the automatic driving vehicle; δ is the acceleration index, and the parameter is 1.86; s(v SV (k), T) is a function; Δv(k) is the relative speed of the automatic driving vehicle and the front vehicle at time k; T is the desired headway of the automatic driving vehicle; b IDMΔt is the unit step length of the trajectory planning of the autonomous vehicle, and the unit is second, and Δy(k) is the longitudinal displacement amount planned for the autonomous vehicle at time k.

[0047] The other steps and parameters are the same as those in Embodiment 1 or 2.

[0048] Embodiment 4: Different from any one of Embodiments 1 to 3, in the step three, the game rules of the game participants are established, the autonomous vehicle is the Leader, which needs to make a decision first, and the rear vehicles in the two lanes are the Followers, which respond to the decision of the Leader (steps 21 to 24);

[0049] Based on the response of the Followers to the decision of the Leader, the optimal decision of the autonomous vehicle that satisfies the constraint condition is calculated according to the reverse induction method, and the lateral displacement amount of the autonomous vehicle at the next time is determined based on the optimal decision.

[0050] The behavior decision of the autonomous vehicle is divided into lateral displacement amount decision and longitudinal displacement amount decision, the lateral displacement amount is determined according to the improved Stackelberg game model, and the longitudinal displacement amount is determined according to the IDM model.

[0051] According to the improved Stackelberg game model, the autonomous vehicle first needs to traverse the strategy set integrated in the vehicle-mounted computing unit to obtain the execution strategy at the next time; secondly, the surrounding vehicles that will conflict with the autonomous vehicle in the driving environment under different execution decisions are determined, and the autonomous vehicle and these vehicles jointly serve as the game participants; finally, according to the Stackelberg game process, the autonomous vehicle predicts the optimal response of other game participants for different execution decisions of the autonomous vehicle, and based on this, makes the optimal decision that maximizes its own benefit, and outputs the lateral displacement amount of the autonomous vehicle at the next time.

[0052] When the autonomous vehicle drives on the highway, the speed of the front vehicle is too slow or the headway is too small, and it is difficult to achieve the expected driving speed or headway. Under the condition of safe driving, the driving environment can be improved by changing lanes. When the autonomous vehicle changes lanes, there will be a conflict of interest with the rear vehicle of the target lane, and the rear vehicle of the target lane can allow or prevent the autonomous vehicle to complete the lane change by changing the acceleration of the vehicle. Therefore, the game participants in the lane changing process of the autonomous vehicle are the vehicle itself and the rear vehicles (FV1, FV2) in the two lanes. The game participants of the autonomous vehicle in the process of changing lanes are N, the strategy set of the autonomous vehicle is A SV , and the strategy set of other game participants is A FV .

[0053] The specific process is:

[0054] Step 3.1. Determine the game participants, which include the autonomous vehicle and the vehicles closest to it in the lanes on either side of the autonomous vehicle.

[0055] N∈{SV,FV1,FV2}, where SV represents the autonomous vehicle, FV1 represents the vehicle closest to the autonomous vehicle in the left lane, and FV2 represents the vehicle closest to the autonomous vehicle in the right lane.

[0056] N represents SV, FV1 or FV2 among the game participants;

[0057] Step 32: Establish the strategy space of the game participants;

[0058] Step 3: Establish constraints to ensure safe driving under different decisions of the autonomous vehicle;

[0059] Step 34: Establish a revenue function for the autonomous driving vehicle based on Step 31, Step 32, and Step 33;

[0060] Step 3.5: Based on the revenue function of the autonomous vehicle, use the backward induction method to calculate the optimal decision of the autonomous vehicle to meet the constraints. Based on optimal decision Determine the lateral displacement Δx of the autonomous driving vehicle at the next moment.

[0061] The other steps and parameters are the same as those in the first to third embodiments.

[0062] Specific implementation method five: This implementation method differs from specific implementation methods one to four in that: in step three-two, the strategy space of the game participants is established;

[0063] The strategy space is:

[0064] A SV ={TL,KL,TR}

[0065]

[0066] Among them, A SV is the strategy set of the SV among the game participants, TL means that the autonomous vehicle SV will change lanes to the left, KL means that the autonomous vehicle SV will stay in the current lane, and TR means that the autonomous vehicle SV will change lanes to the right;

[0067] is the strategy set of FV1 among the game participants, is the jerk of FV1 at time k, the strategy set of FV2 in the game participants, the jerk of FV2 at time k;

[0068] the acceleration of FV1 at time k; the acceleration of FV1 at time k-1;

[0069] the acceleration of FV2 at time k; the acceleration of FV2 at time k-1;

[0070] Δt is the trajectory planning step of the autonomous vehicle; SV the execution decision of the autonomous vehicle SV.

[0071] The other steps and parameters are the same as one of the first to fourth embodiments.

[0072] The sixth embodiment is different from one of the first to fifth embodiments in that the step three establishes the constraint condition for safe driving of the autonomous vehicle under different decisions of the autonomous vehicle;

[0073] The specific process is as follows:

[0074] The constraint condition for safe driving of the autonomous vehicle includes the constraint condition for avoiding collision between the autonomous vehicle and surrounding vehicles and the constraint condition for avoiding collision between the autonomous vehicle and the road shoulder;

[0075] if a SV = TL

[0076]

[0077]

[0078] 0.5×w SV +Δx mean <d SV,1

[0079] wherein, the longitudinal relative distance between SV and LV1 at time k, the speed of LV1 at time k, v SV (k) is the speed of the autonomous vehicle SV at time k, Δt represents the trajectory planning step, Δy min the minimum longitudinal relative distance;

[0080] v max is the maximum driving speed of the autonomous vehicle, which is 33.33 m / s; represents the longitudinal relative distance between SV and FV1 at time k;

[0081] wSV is the width of the autonomous vehicle, d SV,1 The distance between the autonomous vehicle and the left boundary of the innermost lane in its direction of travel (the boundary in the upbound and downbound directions);

[0082] If a SV =KL

[0083]

[0084] in, Indicates the longitudinal relative distance between SV and LV2;

[0085] If a SV =TR

[0086]

[0087]

[0088] 0.5×w SV +Δx mean <d SV,2

[0089] Among them, d SV,2 The distance between the autonomous vehicle and the right edge of the outermost lane in its direction of travel; is the longitudinal relative distance between SV and LV3 at time k, is the speed of LV3 at time k; Represents the longitudinal relative distance between SV and FV3 at time k, obtained in step 1.

[0090] The other steps and parameters are the same as those in the first to fifth embodiments.

[0091] Specific embodiment seven: This embodiment differs from any one of specific embodiments one to six in that: in step three-four, the autonomous driving vehicle benefit function is established based on step three-one, step three-two, and step three-three; the specific process is:

[0092] Constructing a penalty function f for autonomous vehicles SV And benefit function. The benefit function of the autonomous driving vehicle under various driving conditions after executing different decisions is divided into safety benefit function U safe and spatial benefit function U spaceThe safety benefit function, represented by the longitudinal relative distance between SV and FV after t seconds, assesses the risk of rear-end collisions with the vehicle behind the autonomous vehicle after completing a lane change. The spatial benefit function characterizes the spatial lead benefit achieved by the autonomous vehicle after executing different driving maneuvers. The necessity of executing a specific decision at the current moment can be quantified using a penalty function, which is characterized by the difference between the vehicle's desired speed and the speed of the vehicle ahead in the target lane, as well as the decisions of other game participants. The smaller the speed difference and jerk, the greater the improvement in driving conditions achieved by executing the decision. While ensuring safe driving, the vehicle becomes increasingly more compelled to execute the decision. The final benefit is characterized by both the single benefit function and the penalty function.

[0093] If a SV =TL

[0094]

[0095] If a SV =KL

[0096]

[0097] U safe =1

[0098]

[0099] If a SV =TR

[0100]

[0101] Among them, U total is the total revenue function, f SV is the penalty function, U safe is the safety benefit function, U space is the spatial benefit function; Constraint is True indicates that the decision of the autonomous driving vehicle meets the safe driving constraint conditions; Constraint is False indicates that the decision of the autonomous driving vehicle does not meet the safe driving constraint conditions;

[0102] w1, w2, w3 and w4 are control parameters;

[0103] T headway is the headway;

[0104] represents the longitudinal relative distance between SV and LV1 at time k, obtained by step 1;

[0105] v SV (k) represents the speed of the autonomous vehicle SV at time k; represents the velocity of FV1 at time k; represents the velocity of FV2 at time k;

[0106] v des is the expected speed of the autonomous vehicle, is the speed of LV1 at time k;

[0107] is the jerk of FV1 at time k;

[0108] represents the longitudinal relative distance between SV and FV1 at time k, obtained by step 1;

[0109] represents the longitudinal relative distance between SV and FV1 at time k+1;

[0110] y SV (k+1) and It is an intermediate variable and does not need to measure the specific value;

[0111] y SV (k) and Indicates an intermediate variable, and no specific value needs to be measured;

[0112] acc SV (k) is the acceleration of the SV car at time k, and its initial value is set to 0. is the acceleration of FV1 car at time k, and its initial value is set to 0;

[0113] Δt is the unit step length of the autonomous vehicle trajectory planning;

[0114] represents the longitudinal relative distance between SV and LV2 at time k, obtained in step 1;

[0115] is the speed of LV2 at time k;

[0116] represents the longitudinal relative distance between SV and FV2 at time k, obtained by step 1;

[0117] represents the longitudinal relative distance between SV and FV2 at time k+1;

[0118] represents the longitudinal relative distance between SV and LV3 at time k, obtained in step 1;

[0119] is the speed of LV3 at time k;

[0120] is the jerk of FV2 at time k;

[0121] Indicates an intermediate variable, and no specific value needs to be measured;

[0122] represents the velocity of FV2 at time k;

[0123] is the acceleration of FV2 car at time k, and its initial value is set to 0;

[0124] is the jerk of FV2 at time k.

[0125] The other steps and parameters are the same as those in the first to sixth embodiments.

[0126] Specific embodiment eight: This embodiment differs from any one of specific embodiments one to seven in that: in steps three and five, based on the autonomous driving vehicle's benefit function, the optimal decision of the autonomous driving vehicle to satisfy the constraints is calculated according to the reverse induction method. Based on optimal decision Determine the lateral displacement Δx of the autonomous vehicle at the next moment;

[0127] The specific process is:

[0128] According to the backward induction method, the optimal decision of the autonomous driving vehicle to meet the constraints is calculated Determine the lateral displacement Δx of the autonomous vehicle:

[0129]

[0130] in, For the optimal decision of autonomous vehicles, is the strategy candidate set of FV1 based on SV decision, is the strategy candidate set of FV2 based on SV decision, Decision-making for the execution of FV1; Makes decisions about the execution of FV2;

[0131] ξ is the response of FV1 based on the decision of the autonomous vehicle;

[0132] ζ is the response of FV2 based on the decision of the autonomous vehicle;

[0133] Δx is the lateral displacement of the autonomous vehicle SV at the next moment, Δx mean is the average lateral displacement of the autonomous driving vehicle SV per unit step length; is the optimal decision for the autonomous driving vehicle SV.

[0134] The other steps and parameters are the same as those in the first to seventh embodiments.

[0135] Specific embodiment nine: This embodiment is a computer storage medium, which stores at least one instruction. The at least one instruction is loaded and executed by a processor to implement the trajectory planning method during the lane change process of an autonomous driving vehicle.

[0136] It should be understood that the instructions include computer program products, software, or computerized methods corresponding to any method described in the present invention; the instructions can be used to program a computer system or other electronic device. Computer storage media may include readable media on which instructions are stored, and may include but are not limited to magnetic storage media, optical storage media; magneto-optical storage media include read-only memory ROM, random access memory RAM, erasable programmable memory (e.g., EPROM and EEPROM) and flash memory layers, or other types of media suitable for storing electronic instructions.

[0137] Specific embodiment 10: This embodiment is a trajectory planning device for an autonomous vehicle during lane change. The device includes a processor and a memory. It should be understood that the device includes any device including a processor and a memory described in the present invention. The device may also include other units or modules that perform display, interaction, processing, control, and other functions through signals or instructions.

[0138] At least one instruction is stored in the memory, and the at least one instruction is loaded and executed by the processor to implement the trajectory planning method during the lane changing process of an autonomous driving vehicle.

[0139] Those skilled in the art will appreciate that at least one instruction stored is a computer program product corresponding to the method or system. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and interpreted scripting language JavaScript, etc.

[0140] The present application is described with reference to the flowcharts and / or block diagrams of the methods, systems, and computer program products according to the embodiments of the present application, and can also be used for corresponding devices. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0141] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0143] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0144] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

[0145] The present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A trajectory planning method for an autonomous vehicle during lane change, characterized by: The specific process of the method is: Step 1: Get an autonomous vehicle Ego-motion information and autonomous vehicles Geometric dimensions; Get the vehicles around the autonomous driving vehicle The geometric dimensions of the autonomous vehicle and the surrounding vehicles Movement status information; Step 2: Based on the autonomous driving vehicle's own motion information from step 1 and the geometric dimensions and motion state information of vehicles surrounding the autonomous driving vehicle, a longitudinal displacement decision is made; Step 3: Establish the game rules for the game participants. The autonomous vehicle is the leader and needs to make the first decision. The vehicles behind it in the lanes on both sides are followers and respond based on the leader's decision. Based on the Follower's response to the Leader's decision, the optimal decision of the autonomous vehicle that satisfies the constraints is calculated using the backward induction method. The lateral displacement of the autonomous vehicle at the next moment is determined based on the optimal decision. Step 4: Determine the trajectory sequence points of the autonomous driving vehicle at the next moment based on the lateral displacement and longitudinal displacement; Step 5: Set the total lane change time to L seconds and repeat steps 1 to 4. times, and obtain the corresponding total revenue of the autonomous driving vehicle , select the total revenue of the autonomous driving vehicle The lane corresponding to the maximum value is the target lane. The autonomous vehicle executes the optimal decision each time to achieve the final lane-changing trajectory of the autonomous vehicle. The unit step size of the autonomous vehicle trajectory planning, in seconds; In step 3, the game rules for the game participants are established. The autonomous vehicle is the leader and needs to make the decision first. The vehicles behind it in the lanes on both sides are followers and respond according to the leader's decision. Based on the Follower's response to the Leader's decision, the optimal decision of the autonomous vehicle that satisfies the constraints is calculated using the backward induction method. The lateral displacement of the autonomous vehicle at the next moment is determined based on the optimal decision. The specific process is: Step 3.

1. Determine the game participants, which include the autonomous vehicle and the vehicles closest to it in the lanes on either side of the autonomous vehicle. , represents an autonomous vehicle, Indicates the vehicle closest to the autonomous vehicle in the left lane of the autonomous vehicle; Indicates the vehicle closest to the autonomous vehicle in the lane to the right of the autonomous vehicle; Indicates the number of game participants 、 or ; Step 32: Establish the strategy space of the game participants; Step 3: Establish constraints to ensure safe driving under different decisions of the autonomous vehicle; Step 34: Establish a revenue function for the autonomous driving vehicle based on Step 31, Step 32, and Step 33; Step 3.5: Based on the revenue function of the autonomous vehicle, use the backward induction method to calculate the optimal decision of the autonomous vehicle to meet the constraints. , based on the optimal decision Determine the lateral displacement of the autonomous vehicle at the next moment ; In step 33, constraints are established to ensure safe driving under different decisions of the autonomous driving vehicle; The specific process is: The safe driving constraints of the autonomous vehicle include constraints for avoiding collisions between the autonomous vehicle and surrounding vehicles and avoiding collisions between the autonomous vehicle and the roadside; like in, for time and The longitudinal relative distance, for time speed, For autonomous vehicles exist The speed of time, represents the trajectory planning step length, is the minimum longitudinal relative distance; is the maximum speed of the autonomous vehicle, which is 33.33 m / s; express time and The longitudinal relative distance; is the width of the autonomous vehicle, The distance between the autonomous vehicle and the left boundary of the innermost lane in the direction of travel; like in, express and The longitudinal relative distance; like in, The distance between the autonomous vehicle and the right edge of the outermost lane in the direction of travel; for time and The longitudinal relative distance, for time speed; express and exist The vertical relative distance at each moment; In step 34, the autonomous driving vehicle profit function is established based on step 31, step 32, and step 33. The specific process is as follows: like like like in, is the total revenue function, is the penalty function, is the security benefit function, is the spatial benefit function; Indicates that the decision of the autonomous vehicle satisfies the safe driving constraints; Indicates that the decision of the autonomous vehicle does not meet the safe driving constraints; 、 、 and is the control parameter; is the headway; express and exist The vertical relative distance at each moment; Represents autonomous vehicles exist the speed of the moment; express exist the speed of the moment; express exist the speed of the moment; is the expected speed of the autonomous vehicle, for exist the speed of the moment; for time jerk; express and exist The vertical relative distance at each moment; express and exist The vertical relative distance at each moment; and It is an intermediate variable and does not need to measure the specific value; and Represents an intermediate variable, and no specific value needs to be measured; for Car in The acceleration at the moment, the initial value is set to 0, for Car in The acceleration at the moment, the initial value is set to 0; Plan unit step lengths for autonomous vehicle trajectories; express and exist The vertical relative distance at each moment; for exist the speed of the moment; express and exist The vertical relative distance at each moment; express and exist The vertical relative distance at each moment; express and exist The vertical relative distance at each moment; for exist the speed of the moment; for time jerk; 、 Represents an intermediate variable, and no specific value needs to be measured; express exist the speed of the moment; for Car in The acceleration at the moment, the initial value is set to 0; for time acceleration.

2. The trajectory planning method for an autonomous driving vehicle during lane change according to claim 1, characterized in that: In step 1, the autonomous driving vehicle is obtained Ego-motion information and autonomous vehicles Geometric dimensions; Get the vehicles around the autonomous driving vehicle The geometric dimensions of the autonomous vehicle and the surrounding vehicles Movement status information; The specific process is: Autonomous vehicles obtain autonomous vehicle data through on-board computing units. Ego-motion information and autonomous vehicles Geometric dimensions; autonomous vehicles Self-motion information including speed , acceleration ; autonomous vehicles Geometric dimensions including vehicle length , vehicle width ; The autonomous vehicle obtains information about the vehicles around it through the on-board detection unit. The geometric dimensions of the autonomous vehicle and the surrounding vehicles Movement status information; Vehicles surrounding the autonomous vehicle The geometric dimensions include vehicle length , vehicle width ; Vehicles surrounding the autonomous vehicle The motion state information includes speed , acceleration , autonomous driving vehicles and surrounding vehicles The longitudinal relative distance , the distance between the autonomous vehicle and the left boundary of the innermost lane in the direction of travel , the distance between the autonomous vehicle and the right boundary of the outermost lane in the direction of travel ; in, , Indicates the vehicle in front of the autonomous vehicle that is closest to the autonomous vehicle in the left lane; Indicates the vehicle in front of the autonomous vehicle that is closest to the autonomous vehicle in its current lane; Indicates the vehicle in the lane to the right of the autonomous vehicle that is closest to the autonomous vehicle ahead; Indicates the vehicle closest to the autonomous vehicle in the left lane of the autonomous vehicle; Indicates the vehicle behind the autonomous vehicle that is closest to the autonomous vehicle in the lane to the right of the autonomous vehicle.

3. The trajectory planning method for an autonomous driving vehicle during lane change according to claim 2, characterized in that: In step 2, a longitudinal displacement decision of the autonomous driving vehicle is made based on the autonomous driving vehicle's own motion information and the geometric dimensions and motion state information of vehicles surrounding the autonomous driving vehicle obtained in step 1; The specific process is: Calculation of the longitudinal displacement of an autonomous driving vehicle based on the IDM model ; The IDM model is: in, express The acceleration of the autonomous driving vehicle at all times, with the initial value set to 0; is the maximum acceleration of the autonomous vehicle; for The speed of the autonomous driving vehicle at all times, the initial value is set to 0; for time speed; is the expected speed of the autonomous vehicle; is the acceleration index; is a function; for The relative speed between the autonomous driving vehicle and the vehicle in front at any moment, is the expected headway of the autonomous vehicle; is the maximum deceleration of the autonomous vehicle, is the unit step length of the autonomous vehicle trajectory planning, in seconds, for The time is the longitudinal displacement planned by the autonomous driving vehicle, for The distance between the autonomous driving vehicle and the vehicle in front at all times, is the minimum desired vehicle distance.

4. The trajectory planning method for an autonomous driving vehicle during lane change according to claim 3, characterized in that: In step 32, the strategy space of the game participants is established; The strategy space is: in, For game participants The strategy set, Represents autonomous vehicles Will change lanes to the left, Represents autonomous vehicles Will maintain current lane, Represents autonomous vehicles Will change lanes to the right; For game participants The strategy set, for time The acceleration of For game participants The strategy set, for time jerk; for Car in acceleration of the moment; for Car in acceleration of the moment; for Car in acceleration of the moment; for Car in acceleration of the moment; Plan the step size for autonomous vehicle trajectories; For autonomous vehicles implementation decisions.

5. The trajectory planning method for an autonomous driving vehicle during lane change according to claim 4, characterized in that: In steps 3 and 5, based on the autonomous driving vehicle's profit function, the optimal decision of the autonomous driving vehicle to meet the constraints is calculated according to the backward induction method. , based on the optimal decision Determine the lateral displacement of the autonomous vehicle at the next moment ; The specific process is: According to the backward induction method, the optimal decision of the autonomous driving vehicle to meet the constraints is calculated , determine the lateral displacement of the autonomous vehicle : in, For the optimal decision of autonomous vehicles, for based on The set of candidate strategies for decision making, for based on The set of candidate strategies for decision making, for implementation decisions; for implementation decisions; for Responses based on autonomous vehicle decisions; for Responses based on autonomous vehicle decisions; For autonomous vehicles The lateral displacement at the next moment, For autonomous vehicles The average lateral displacement per unit step length; For autonomous vehicles The optimal decision.

6. A computer storage medium, characterized in that: The storage medium stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement a trajectory planning method for an autonomous driving vehicle during lane changing as described in any one of 1 to 5.

7. A trajectory planning device for an autonomous vehicle during lane change, characterized in that: The device includes a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement a trajectory planning method for an autonomous driving vehicle during a lane change process as described in any one of claims 1 to 5.

Images