Vehicle lane changing control method and device, electronic equipment and storage medium
By tracking the associated vehicle in the bicycle lane change mode and establishing a lane change trajectory planning model that takes into account the characteristics of random motion, using deep reinforcement learning optimization control strategy, the safety caused by the uncertainty of the associated vehicle during lane change is solved. Sex and comfort issues achieve more efficient lane change control.
Patent Information
- Application Number
- CN202411024857.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-15
AI Technical Summary
During the active lane change of the bicycle, the random motion characteristics of the associated vehicle lead to a large deviation from the actual lane change process and the predicted trajectory, affecting the safety and comfort of the bicycle.
When the bicycle enters lane change mode, track and determine the associated vehicle, and establish a lane change trajectory planning model based on its driving data, considering the random motion characteristics of the associated vehicle, and optimizing the lane change control strategy through a deep reinforcement learning algorithm to reduce driving risk.
It effectively improves the safety and comfort of the bicycle lane change process, and reduces the driving risks during lane change by accurately controlling the impact of related vehicle motion uncertainty.
Smart Images

Figure CN118953352A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent driving technology, and in particular to a vehicle lane change control method, device, electronic device and storage medium. Background Art
[0002] At present, active lane change function has become a necessary function of high-level autonomous driving technology. The safety, smoothness and comfort of active lane change process are important indicators for evaluating the intelligence level of smart cars. However, during the active lane change process of the ego vehicle, the motion trajectory of the associated vehicles that affect the driving of the ego vehicle is uncertain. Its random motion characteristics lead to a large deviation between the actual lane change process of the ego vehicle and the predicted trajectory, thus affecting the safety and comfort of the active lane change process of the ego vehicle. Summary of the invention
[0003] The present application discloses a vehicle lane change control method, device, electronic device and storage medium to solve the problems of vehicle lane change safety and comfort caused by the random motion characteristics of associated vehicles.
[0004] In a first aspect, an embodiment of the present application provides a vehicle lane change control method, the method comprising:
[0005] When the ego vehicle enters the lane change mode, the associated vehicle is tracked and determined when the ego vehicle changes lanes, and the associated vehicle represents the vehicle that has an impact on the lane change of the ego vehicle;
[0006] Based on the driving data of the associated vehicles, a lane change trajectory planning model corresponding to the ego vehicle is established. The lane change trajectory planning model takes into account the random motion characteristics of the associated vehicles.
[0007] Based on the lane-changing trajectory planning model, the lane-changing process of the vehicle is controlled.
[0008] In a second aspect, an embodiment of the present application provides a vehicle lane change control device, the device comprising:
[0009] A tracking module is used to track and determine the associated vehicle when the ego vehicle changes lanes when the ego vehicle enters the lane change mode, and the associated vehicle represents the vehicle that has an impact on the lane change of the ego vehicle;
[0010] A planning module, for establishing a lane change trajectory planning model corresponding to the ego vehicle based on the driving data of the associated vehicle, wherein the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle;
[0011] A control module is used to control the lane changing process of the vehicle based on the lane changing trajectory planning model.
[0012] Optionally, the tracking module is used to:
[0013] When the ego vehicle enters the lane change mode, the current lane of the ego vehicle and the target lane of the ego vehicle's lane change are determined;
[0014] In the current lane, a first front vehicle located in front of the vehicle and a first rear vehicle located behind the vehicle are determined, and in the target lane, a second front vehicle located in front of the vehicle and a second rear vehicle located behind the vehicle are determined;
[0015] The first front vehicle, the first rear vehicle, the second front vehicle, and the second rear vehicle are regarded as associated vehicles.
[0016] Optionally, the planning module is used to:
[0017] Obtaining driving data of the associated vehicle, the driving data including position, speed and acceleration information;
[0018] Based on the driving data, a first sub-planning model for changing lanes in a first direction and a second sub-planning model for changing lanes in a second direction are established. The first direction is the driving direction of the vehicle when it moves parallel to the target lane, and the second direction is the forward direction of the vehicle perpendicular to the first direction.
[0019] The first sub-planning model and the second sub-planning model are combined to obtain a lane-changing trajectory planning model, which is a Gaussian process model that takes into account the random motion characteristics of associated vehicles.
[0020] Optionally, the control module is used to:
[0021] Based on the lane change trajectory planning model, determine the desired position of the vehicle;
[0022] Compare the expected posture with the current posture of the vehicle to obtain the posture deviation; the posture deviation includes the longitudinal displacement deviation, lateral displacement deviation and lane change orientation deviation of the vehicle;
[0023] Based on the posture deviation, the lane changing process of the vehicle is controlled.
[0024] Optionally, the control module is used to:
[0025] Based on the posture deviation, an active lane change control reward function is established that takes into account the random motion characteristics of the associated vehicles.
[0026] Based on the active lane change control reward function, an active lane change control model is established; the active lane change control model is trained using the deep reinforcement learning PPO algorithm to obtain the optimal control strategy;
[0027] Based on the optimal control strategy, the lane changing process of the vehicle is controlled.
[0028] Optionally, the control module is used to:
[0029] Based on the posture deviation, the trajectory tracking index function corresponding to the ego vehicle is established;
[0030] Based on the acceleration change of the vehicle in two adjacent control cycles, a comfort index function is established;
[0031] Based on the trajectory tracking index function and the comfort index function, the active lane changing control reward function considering the random motion characteristics of the associated vehicles is established by utilizing the Markov property of the lane changing process of the ego vehicle.
[0032] Optionally, the PPO algorithm includes a Critic network and an Actor network, and the control module is used to:
[0033] The active lane-changing control model corresponding to the lane-changing trajectory planning model is evaluated based on the critic network to obtain the evaluation results;
[0034] Based on the evaluation results, the active lane change control model is improved using the Actor network to obtain an improved control strategy.
[0035] Based on the action decision distribution ratio between the active lane change control model and the improved control strategy, the optimal control strategy is obtained.
[0036] In a third aspect, the present application provides an electronic device, including:
[0037] Memory, used to store computer programs;
[0038] The processor is used to implement the above-mentioned vehicle lane change control method when executing the computer program stored in the memory.
[0039] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned vehicle lane change control method is implemented.
[0040] The beneficial effects of this application are as follows:
[0041] The present application provides a vehicle lane change control method, device, electronic device and storage medium. When the vehicle enters the lane change mode, the associated vehicle is tracked and determined when the vehicle changes lanes, and based on the driving data of the associated vehicle, a lane change trajectory planning model corresponding to the vehicle is established. Based on the lane change trajectory planning model, the lane change process of the vehicle is controlled. Among them, the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle. In this way, the lateral and longitudinal control problems of the active lane change of the vehicle under the condition of the uncertainty of the associated vehicle motion are solved, and the driving risk during the lane change process of the vehicle caused by the random motion characteristics of the associated vehicle is reduced, which can effectively improve the safety and comfort during the lane change process. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 A flow chart of a vehicle lane change control method provided in this application;
[0043] Figure 2 A schematic diagram of a vehicle lane changing scenario provided for this application;
[0044] Figure 3 A flow chart of a vehicle lane change control method provided in this application;
[0045] Figure 4 A flowchart of a PPO model training method provided in this application;
[0046] Figure 5 A schematic diagram of the structure of a vehicle lane change control device provided in this application;
[0047] Figure 6 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION
[0048] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The specific operating methods in the method embodiments can also be applied to device embodiments or system embodiments. It should be noted that in the description of the present application, "multiple" is understood as "at least two". "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A is connected to B, which can represent: A is directly connected to B and A is connected to B through C. In addition, in the description of the present application, words such as "first" and "second" are only used to distinguish the purpose of description, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order.
[0049] The embodiments of the present application are described in detail below in conjunction with the accompanying drawings.
[0050] At present, active lane change function has become a necessary function of high-level autonomous driving technology. The safety, smoothness and comfort of active lane change process are important indicators for evaluating the intelligence level of smart cars. However, during the active lane change process of the ego vehicle, the motion trajectory of the associated vehicles that affect the driving state of the ego vehicle is uncertain. Its random motion characteristics lead to a large deviation between the actual lane change process of the ego vehicle and the predicted trajectory, thus affecting the safety, comfort and efficiency of the active lane change process of the ego vehicle.
[0051] In order to solve the above problems, the present application provides a vehicle lane change control method, device, electronic device and storage medium. When the vehicle enters the lane change mode, the associated vehicle when the vehicle changes lanes is tracked and determined, and based on the driving data of the associated vehicle, a lane change trajectory planning model corresponding to the vehicle is established. Based on the lane change trajectory planning model, the lane change process of the vehicle is controlled. Among them, the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle. In this way, the lateral and longitudinal control problems of the vehicle's active lane change under the uncertain conditions of the associated vehicle's motion are solved, and the driving risk during the lane change process of the vehicle caused by the random motion characteristics of the associated vehicle is reduced, which can effectively improve the safety and comfort during the lane change process. It should be noted that the method, device, electronic device and storage medium in the embodiments of the present application are based on the same technical concept, and the principles of solving the problems are similar. Therefore, the embodiments of the device, electronic device and storage medium and the method can refer to each other, and the repeated parts will not be repeated.
[0052] In particular, the preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application, and the embodiments of the present application and the features in the embodiments may be combined with each other if there is no conflict.
[0053] like Figure 1 FIG. 1 is a flow chart of a vehicle lane change control method provided by the present application, which specifically includes the following steps:
[0054] S101, when the vehicle enters a lane change mode, tracking and determining associated vehicles when the vehicle changes lanes;
[0055] In the embodiment of the present application, the vehicle can start the lane change mode in a variety of ways, such as by system active triggering, button triggering, or voice activation. Among them, the button triggering includes virtual buttons or physical buttons, which are not limited here. When the vehicle enters the lane change mode, it indicates that the vehicle has started the lane change preparation, which does not mean that the lane change operation will be performed immediately.
[0056] When the ego vehicle enters the lane change mode, it is necessary to track and determine the associated vehicles when the ego vehicle changes lanes. The associated vehicles represent vehicles that may affect the lane change of the ego vehicle. The method for determining the associated vehicles includes:
[0057] When the vehicle enters the lane change mode, the current lane of the vehicle and the target lane of the vehicle are tracked and determined; in the current lane, the first front vehicle in front of the vehicle and the first rear vehicle behind the vehicle are determined, and in the target lane, the second front vehicle in front of the vehicle and the second rear vehicle behind the vehicle are determined; the first front vehicle, the first rear vehicle, the second front vehicle, and the second rear vehicle are regarded as associated vehicles.
[0058] In a possible application scenario, the associated vehicles of the self-vehicle also include:
[0059] At the start of the lane change, a first side vehicle of the vehicle is located in a non-target lane adjacent to the current lane of the vehicle, and a second side vehicle of the vehicle is located in a non-current lane adjacent to the target lane after the vehicle enters the target lane.
[0060] For example, refer to Figure 2 , is a schematic diagram of the vehicle lane change application scenario. Figure 2 In the example, there are 4 lanes, among which, from top to bottom, the 3rd lane is the current lane, the 2nd lane is the target lane, the 4th lane is the non-target lane adjacent to the current lane, and the 1st lane is the non-current lane adjacent to the target lane. Then, vehicle Ego is the ego vehicle, vehicle Fo is the first front vehicle, vehicle Ro is the first rear vehicle, vehicle F1 is the second front vehicle, vehicle R1 is the second rear vehicle, vehicle TV-R is the first side vehicle, and vehicle TV-L is the second side vehicle.
[0061] Through the above method, the associated vehicles that have an impact on the lane change of the own vehicle can be determined.
[0062] It should be noted that the determination of associated vehicles is related to the number of lanes and the positions and speeds of vehicles in each lane, and may vary depending on the specific circumstances. As long as there is a possibility of affecting the lane change of the vehicle itself, it is an associated vehicle, and the application scenarios will not be exhaustively listed here for explanation.
[0063] S102, based on the driving data of the associated vehicle, establishing a lane change trajectory planning model corresponding to the ego vehicle, wherein the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle;
[0064] After the associated vehicle is determined, the driving data of the associated vehicle is further obtained, wherein the driving data includes position, speed and acceleration information.
[0065] Furthermore, based on the driving data, a lane change trajectory planning model corresponding to the vehicle is established.
[0066] The lane change trajectory planning model is a polynomial function f(x, y, t) that constrains the vehicle's travel, where x represents the first direction, y represents the second direction, and t represents the time.
[0067] After decomposing the polynomial function f(x,y,t) into polynomials in the x-direction and y-direction, the resulting model is as follows:
[0068]
[0069] Among them, a0~a n , b0~b n is the fitting coefficient;
[0070] is the state quantity of the vehicle at the initial lane change moment;
[0071] is the state quantity at the end of the lane change of the vehicle;
[0072] x begin , is the position, velocity and acceleration in the x direction at the initial lane change moment;
[0073] x end , is the position, velocity and acceleration in the x direction at the time of target lane change (end);
[0074] y begin , is the position, velocity and acceleration in the y direction at the initial lane change moment;
[0075] y end , are the position, velocity and acceleration in the y direction at the target lane change (end) moment.
[0076] The above process includes the following three aspects:
[0077] First, based on the driving data, a first sub-planning model for changing lanes of the vehicle in a first direction is established, wherein the first direction is the driving direction of the vehicle when it moves parallel to the target lane.
[0078] For example, see Figure 2 At the initial lane change moment, the longitudinal distance L between the ego vehicle Ego and the vehicle Fo in the x direction is Fo , the vehicle speed is V Fox , relative to the longitudinal distance L of the vehicle Ro Ro , the vehicle speed is V Rox At the end of the lane change, the longitudinal distance L between the vehicle Ego and the vehicle F1 in the x direction is F1 and vehicle speed V F1x , relative to the longitudinal distance L of vehicle R1 R1 , the vehicle speed is V R1x , then, the first sub-programming model established can be:
[0079]
[0080] Second, based on the driving data, a second sub-planning model for the vehicle to change lanes in a second direction is established, wherein the second direction is a forward direction of the vehicle that is perpendicular to the first direction.
[0081] For example, see Figure 2 The application scenario diagram shown.
[0082] exist Figure 2 In the initial lane change moment, the lateral distance between the ego vehicle Ego and the vehicle R1 in the y direction is D R1 , the vehicle speed is V R1y , lateral distance D relative to vehicle TV-R TV-R , the vehicle speed is V TV-Ry ; At the end of the lane change, the lateral distance D between the vehicle Ego and the vehicle TV-L in the y direction is TV-L , the vehicle speed is V TV-Ly , the lateral distance D relative to the vehicle Fo Fo , the vehicle speed is V Foy .
[0083] Then, the second sub-planning model established can be:
[0084]
[0085] Third, the first sub-planning model and the second sub-planning model are combined.
[0086] In this way, through the above method, a lane change trajectory planning model of the vehicle is obtained, and the lane change trajectory planning model is a Gaussian process model that takes into account the random motion characteristics of the associated vehicle.
[0087] S103, controlling the lane changing process of the vehicle based on the lane changing trajectory planning model.
[0088] In the embodiment of the present application, after establishing the lane change trajectory planning model corresponding to the vehicle, further,
[0089] Based on the lane change trajectory planning model, the lane change process of the vehicle is controlled, including:
[0090] Based on the lane change trajectory planning model, the desired posture of the ego vehicle is determined, and the desired posture is compared with the current posture of the ego vehicle to obtain the posture deviation, which includes the longitudinal displacement deviation, lateral displacement deviation, and lane change orientation deviation of the ego vehicle. Furthermore, based on the posture deviation, the lane change process of the ego vehicle is controlled.
[0091] Among them, the mathematical model corresponding to the posture deviation, that is, the kinematic model of the active lane change deviation characteristic of the ego vehicle, can be:
[0092]
[0093] Among them, x δ is the longitudinal position deviation of the vehicle changing lanes, y δ is the lateral position deviation of the vehicle when changing lanes, θ δ is the lane change direction deviation of the vehicle; e ,ω e is the current speed and yaw rate of the vehicle, νc ,ω c are the velocity and yaw rate of the desired vehicle posture.
[0094] In one embodiment, reference Figure 3 The above-mentioned control of the lane change process of the vehicle based on the posture deviation includes the following steps:
[0095] S301, based on the posture deviation, establishing an active lane change control reward function that takes into account the random motion characteristics of the associated vehicle, and establishing an active lane change control model based on the active lane change control reward function.
[0096] The specific implementation of this step is as follows:
[0097] First, based on the posture deviation, the trajectory tracking index function Q1 corresponding to the vehicle is established, Q1 = W1*△ δ , where W1 is the trajectory tracking weight coefficient;
[0098] Then, based on the acceleration change of the vehicle in two adjacent control cycles, the comfort index function Q2 is established;
[0099] Finally, based on the trajectory tracking index function and the comfort index function, the Markov property of the lane changing process of the ego vehicle is utilized to establish an active lane changing control reward function R that takes into account the random motion characteristics of the associated vehicle.
[0100] It should be noted that due to the uncertainty of the speed and distance of the associated vehicles, the active lane change of the ego vehicle is a random process, and the function corresponding to the active lane change control model that controls the ego vehicle is an uncertain function. Therefore, in the embodiment of the present application, assuming that the active lane change process satisfies the Markov property, the active lane change control process that considers the randomness of the associated vehicle movement is regarded as an MDP (S, A, P, R). Among them, S is the motion state of the ego vehicle during the active lane change process; A is the ego vehicle control decision u; P is the state transition probability caused by the uncertainty of the associated vehicle movement; R is the reward function.
[0101] The active lane change control sub-strategy is The lane change control problem of the vehicle can be transformed into a strategy to solve the maximum expected total discounted reward in the infinite control time domain, which is the active lane change control model Specifically:
[0102]
[0103] Among them, the discount factor γ∈(0,1), For the strategy The trajectory distribution under time t, the reward function R t =exp(Q1+Q2).
[0104] S302: Use deep reinforcement learning PPO algorithm to train the active lane change control model to obtain the optimal control strategy.
[0105] Among them, the PPO algorithm includes the Critic network and the Actor network, so the specific implementation method of this step includes:
[0106] The active lane change control model corresponding to the lane change trajectory planning model is evaluated based on the Critic network to obtain an evaluation result; based on the evaluation result, the active lane change control model is improved by using the Actor network to obtain an improved control strategy; based on the action decision distribution ratio between the active lane change control model and the improved control strategy, an optimal control strategy is obtained, including: based on the action decision distribution ratio between the active lane change control model and the improved control strategy, the active lane change control model is updated, and the update amplitude of the control strategy is adjusted based on the control strategy update amplitude value to obtain the optimal control strategy.
[0107] In an application scenario, refer to Figure 2 , this step is implemented as follows:
[0108] An active lane change control architecture for the ego vehicle is established under the random motion state of associated vehicles. The system controller simultaneously takes the ego vehicle Ego, vehicle Fo, vehicle Ro, vehicle F1, vehicle R1, vehicle TV-R, and vehicle TV-L as inputs, and uses the PPO deep reinforcement learning algorithm to establish the optimal lateral and longitudinal control strategy for the ego vehicle Ego.
[0109] The PPO algorithm is a reinforcement learning algorithm based on the Actor-Critic architecture. It uses the Critic network and the Actor network to simultaneously estimate the value function and the optimal strategy of the system. The Critic network is used to estimate the value function to evaluate the current strategy; the Actor network improves the strategy based on the estimated results of the value function to obtain the optimal control strategy of the system.
[0110]
[0111] Among them, V(s t ) is state s t The value function of Indicates that the probability of taking the current action should be increased; Indicates that the probability of taking this action should be reduced.
[0112] The Critic network and the Actor network are updated using the gradient algorithm, and their objective functions are expressed as:
[0113]
[0114] Where: p t (θ)=π θ (a t |s t ) / π π_old (a t |s t ) is the new strategy π in the strategy update process θ and the old strategy π π_old The ratio of action decision distribution; ε is a constant, here the value is ε = 0.2.
[0115] In the learning process, use J actor and J critic The gradient of the Actor and Critic network parameters is learned to continuously update the policy network and obtain the optimal decision for the longitudinal control of the vehicle. When updating the network parameters, the clipping function is used to avoid a large difference between the new and old policies, and the parameter ε is used to control the policy update amplitude.
[0116] S303, based on the optimal control strategy, the lane changing process of the vehicle is controlled.
[0117] After the optimal control strategy for the lane change of the vehicle is obtained through the above steps S301 and S302, the lane change process of the vehicle can be controlled according to the final control strategy.
[0118] It should be noted that the optimal control strategy for a vehicle is not static, but can be adjusted in a timely manner according to the actual driving environment on site.
[0119] In addition, in the implementation of this application, the training method of the PPO model is further disclosed. The specific process can be referred to Figure 4 Specifically:
[0120] S401, initialize the Critic network and the Actor network;
[0121] S402, cyclically acquiring real-time position data (random motion state) of the associated vehicle during the lane change process of the ego vehicle;
[0122] S403, using the lane change trajectory planning model to predict the driving data of the associated vehicle at the next moment;
[0123] S404, using the Actor network to obtain sample data (St, At, Rt), where t is a positive integer.
[0124] S405, the loop ends, and the sample point set (S1, A1, R1), (S2, A2, R2)...(ST, AT, RT) is obtained;
[0125] S406, calculate the advantage function
[0126] S407, calculate the objective function J of the Actor network actor ;
[0127] S408, update Actor network parameters;
[0128] S409, calculate the objective function J of the Critic network critic ;
[0129] S410, update critic network parameters.
[0130] By repeatedly executing the above process, the PPO model can be trained.
[0131] The above content is an embodiment of a vehicle lane change control method for the vehicle for which protection is requested in this application. When the self-vehicle enters the lane change mode, the associated vehicle when the self-vehicle changes lanes is tracked and determined, and based on the driving data of the associated vehicle, a lane change trajectory planning model corresponding to the self-vehicle is established, and the lane change process of the self-vehicle is controlled based on the lane change trajectory planning model. Among them, the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicles. In this way, the active lane change lateral and longitudinal control problems of the self-vehicle under the uncertain conditions of the associated vehicle motion are solved, and the driving risk during the lane change process of the self-vehicle caused by the random motion characteristics of the associated vehicles is reduced, which can effectively improve the safety and comfort during the lane change process.
[0132] Based on the same inventive concept, the embodiment of the present application also provides a vehicle lane change control device, such as Figure 5 FIG. 1 is a schematic diagram of a vehicle lane change control device in the present application, and the device includes:
[0133] A tracking module 501 is used to track and determine the associated vehicles when the ego vehicle changes lanes when the ego vehicle enters the lane change mode, where the associated vehicles represent vehicles that have an impact on the lane change of the ego vehicle;
[0134] A planning module 502, configured to establish a lane change trajectory planning model corresponding to the ego vehicle based on the driving data of the associated vehicle, wherein the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle;
[0135] The control module 503 is used to control the lane changing process of the vehicle based on the lane changing trajectory planning model.
[0136] Optionally, the tracking module 501 is used to:
[0137] When the ego vehicle enters the lane change mode, the current lane of the ego vehicle and the target lane of the ego vehicle's lane change are determined;
[0138] In the current lane, a first front vehicle located in front of the vehicle and a first rear vehicle located behind the vehicle are determined, and in the target lane, a second front vehicle located in front of the vehicle and a second rear vehicle located behind the vehicle are determined;
[0139] The first front vehicle, the first rear vehicle, the second front vehicle, and the second rear vehicle are regarded as associated vehicles.
[0140] Optionally, the planning module 502 is used to:
[0141] Obtaining driving data of the associated vehicle, the driving data including position, speed and acceleration information;
[0142] Based on the driving data, a first sub-planning model for changing lanes in a first direction and a second sub-planning model for changing lanes in a second direction are established. The first direction is the driving direction of the vehicle when it moves parallel to the target lane, and the second direction is the forward direction of the vehicle perpendicular to the first direction.
[0143] The first sub-planning model and the second sub-planning model are combined to obtain a lane-changing trajectory planning model, which is a Gaussian process model that takes into account the random motion characteristics of associated vehicles.
[0144] Optionally, the control module 503 is used to:
[0145] Based on the lane change trajectory planning model, determine the desired position of the vehicle;
[0146] Compare the expected posture with the current posture of the vehicle to obtain the posture deviation; the posture deviation includes the longitudinal displacement deviation, lateral displacement deviation and lane change orientation deviation of the vehicle;
[0147] Based on the posture deviation, the lane changing process of the vehicle is controlled.
[0148] Optionally, the control module 503 is used to:
[0149] Based on the posture deviation, an active lane change control reward function is established that takes into account the random motion characteristics of the associated vehicles.
[0150] Based on the active lane change control reward function, an active lane change control model is established;
[0151] The active lane change control model is trained using the deep reinforcement learning PPO algorithm to obtain the optimal control strategy;
[0152] Based on the optimal control strategy, the lane changing process of the vehicle is controlled.
[0153] Optionally, the control module 503 is used to:
[0154] Based on the posture deviation, the trajectory tracking index function corresponding to the ego vehicle is established;
[0155] Based on the acceleration change of the vehicle in two adjacent control cycles, a comfort index function is established;
[0156] Based on the trajectory tracking index function and the comfort index function, the active lane changing control reward function considering the random motion characteristics of the associated vehicles is established by utilizing the Markov property of the lane changing process of the ego vehicle.
[0157] Optionally, the PPO algorithm includes a Critic network and an Actor network, and the control module 503 is used to:
[0158] The active lane-changing control model corresponding to the lane-changing trajectory planning model is evaluated based on the critic network to obtain the evaluation results;
[0159] Based on the evaluation results, the active lane change control model is improved using the Actor network to obtain an improved control strategy.
[0160] Based on the action decision distribution ratio between the active lane change control model and the improved control strategy, the optimal control strategy is obtained.
[0161] The above content is an embodiment of a vehicle lane change control device for a vehicle for which protection is requested in this application. When the vehicle enters the lane change mode, the associated vehicle when the vehicle changes lanes is tracked and determined, and based on the driving data of the associated vehicle, a lane change trajectory planning model corresponding to the vehicle is established, and the lane change process of the vehicle is controlled based on the lane change trajectory planning model. Among them, the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicles. In this way, the lateral and longitudinal control problems of the vehicle's active lane change under the uncertain conditions of the associated vehicle's motion are solved, and the driving risk during the lane change process of the vehicle caused by the random motion characteristics of the associated vehicles is reduced, which can effectively improve the safety and comfort during the lane change process.
[0162] Based on the same inventive concept, an electronic device is also provided in the embodiment of the present application, and the electronic device can realize the functions of the aforementioned vehicle lane change control method and device, referring to Figure 6 The electronic device includes: a processor 61, which is used to implement the above-mentioned vehicle lane change control method when executing the computer program stored in the memory; and a memory 62, which is used to store the computer program.
[0163] The specific connection medium between the processor 61 and the memory 62 is not limited in the embodiment of the present application. Figure 6 In the example, the processor 61 and the memory 62 are connected via a bus 60. The bus 60 is Figure 6 The connection between other components is shown by bold lines, and the connection between other components is only for schematic illustration and is not intended to be limiting. The bus 60 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6Only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. Alternatively, the processor 61 can also be called a controller, and there is no limitation on the name.
[0164] In the embodiment of the present application, the memory 62 stores instructions that can be executed by the processor 61. The processor 61 can execute the vehicle lane change control method discussed above by executing the instructions stored in the memory 62. The processor 61 can implement Figure 5 The functions of each module in the device shown.
[0165] Among them, the processor 61 is the control center of the device, and can use various interfaces and lines to connect the various parts of the entire control device. By running or executing instructions stored in the memory 62 and calling data stored in the memory 62, the various functions of the device and processing data, the device can be monitored as a whole.
[0166] In a possible design, the processor 61 may include one or more processing units, and the processor 61 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the modem processor may not be integrated into the processor 61. In some embodiments, the processor 61 and the memory 62 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on separate chips.
[0167] The processor 61 may be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and may implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the vehicle lane change control method disclosed in the embodiments of the present application may be directly embodied as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.
[0168] The memory 62 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 62 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 62 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 62 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.
[0169] By programming the processor 61, the code corresponding to the vehicle lane change control method described in the above embodiment can be fixed into the chip, so that the chip can execute the vehicle lane change control method when running. Figure 1 The steps of the vehicle lane change control method of the embodiment shown are as follows: How to design and program the processor 61 is a technique known to those skilled in the art and will not be described in detail here.
[0170] Based on the same inventive concept, an embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the above-mentioned vehicle lane change control method is implemented.
[0171] In some possible implementations, various aspects of the vehicle lane change control method provided in the present application may also be implemented in the form of a program product, which includes a program code. When the program product is run on a device, the program code is used to enable the control device to execute the steps of the vehicle lane change control method according to various exemplary embodiments of the present application described above in this specification.
[0172] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0173] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. 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 generate 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.
[0174] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.
[0175] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0176] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.
Claims
1. A vehicle lane change control method, characterized in that: The method comprises: When the ego vehicle enters a lane change mode, tracking and determining an associated vehicle when the ego vehicle changes lanes, the associated vehicle representing a vehicle that has an impact on the lane change of the ego vehicle; Based on the driving data of the associated vehicle, a lane change trajectory planning model corresponding to the self-vehicle is established, wherein the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle; Based on the lane change trajectory planning model, the lane change process of the vehicle is controlled.
2. The method according to claim 1, characterized in that: When the vehicle enters the lane change mode, determining the associated vehicle when the vehicle changes lanes includes: When the vehicle enters a lane change mode, determining a current lane of the vehicle and a target lane for the vehicle to change lanes; In the current lane, a first front vehicle located in front of the vehicle and a first rear vehicle located behind the vehicle are determined, and in the target lane, a second front vehicle located in front of the vehicle and a second rear vehicle located behind the vehicle are determined; The first front vehicle, the first rear vehicle, the second front vehicle, and the second rear vehicle are regarded as the associated vehicles.
3. The method according to claim 1, characterized in that: The step of establishing a lane change trajectory planning model corresponding to the vehicle based on the driving data of the associated vehicle includes: Acquiring driving data of the associated vehicle, the driving data including position, speed and acceleration information; Based on the driving data, a first sub-planning model for the vehicle to change lanes in a first direction and a second sub-planning model for the vehicle to change lanes in a second direction are established, wherein the first direction is the driving direction of the vehicle when moving parallel to the target lane, and the second direction is the forward direction of the vehicle perpendicular to the first direction; The first sub-planning model and the second sub-planning model are combined to obtain the lane change trajectory planning model, and the lane change trajectory planning model is a Gaussian process model that takes into account the random motion characteristics of the associated vehicle.
4. The method according to claim 1, characterized in that: The controlling the lane changing process of the vehicle based on the lane changing trajectory planning model includes: Determining a desired posture corresponding to the ego-vehicle based on the lane change trajectory planning model; Comparing the desired posture with the current posture of the vehicle to obtain a posture deviation; wherein the posture deviation includes a longitudinal displacement deviation, a lateral displacement deviation, and a lane change orientation deviation of the vehicle; Based on the posture deviation, a lane changing process of the vehicle is controlled.
5. The method according to claim 4, characterized in that: The controlling the lane changing process of the vehicle based on the posture deviation includes: Based on the posture deviation, establishing an active lane change control reward function that takes into account the random motion characteristics of the associated vehicle; Based on the active lane change control reward function, establishing an active lane change control model; The active lane change control model is trained using a deep reinforcement learning PPO algorithm to obtain an optimal control strategy; Based on the optimal control strategy, the lane changing process of the vehicle is controlled.
6. The method according to claim 5, characterized in that The step of establishing an active lane change control reward function taking into account the random motion characteristics of the associated vehicle based on the posture deviation includes: Based on the posture deviation, establishing a trajectory tracking index function corresponding to the ego-vehicle; Establishing a comfort index function based on the acceleration change of the vehicle in two adjacent control cycles; Based on the trajectory tracking index function and the comfort index function, and utilizing the Markov property of the lane changing process of the ego vehicle, an active lane changing control reward function is established that takes into account the random motion characteristics of the associated vehicle.
7. The method according to claim 5, characterized in that: The PPO algorithm includes a Critic network and an Actor network. The active lane change control model is trained using the deep reinforcement learning PPO algorithm to obtain an optimal control strategy, including: Evaluating the active lane change control model corresponding to the lane change trajectory planning model based on the critic network to obtain an evaluation result; Based on the evaluation result, the active lane change control model is improved by using the Actor network to obtain an improved control strategy; Based on the action decision distribution ratio between the active lane change control model and the improved control strategy, the optimal control strategy is obtained.
8. A vehicle lane change control device, characterized in that: The device comprises: A tracking module, used for tracking and determining an associated vehicle when the ego vehicle changes lanes when the ego vehicle enters a lane change mode, wherein the associated vehicle represents a vehicle that has an impact on the lane change of the ego vehicle; A planning module, for establishing a lane change trajectory planning model corresponding to the ego vehicle based on the driving data of the associated vehicle, wherein the lane change trajectory planning model takes into account the random motion characteristics of the associated vehicle; A control module is used to control the lane changing process of the vehicle based on the lane changing trajectory planning model.
9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to implement the method according to any one of claims 1 to 7 when executing a computer program stored in the memory.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.
