A vehicle intervention trajectory planning method, device and electronic equipment

Abstract

Embodiments of the present application disclose a vehicle intervention trajectory planning method and device and an electronic device, and relate to the technical field of vehicle assisted driving, and are invented to improve the environmental adaptability of an intervention trajectory. The method comprises the following steps: obtaining a current state and a target state of a vehicle; determining an optimal parameter value of a to-be-optimized parameter according to a preset genetic algorithm, an initial value of the to-be-optimized parameter of a to-be-planned intervention trajectory, the current state and the target state; the to-be-optimized parameter is a to-be-optimized parameter of a first motion state parameter model in the to-be-planned intervention trajectory, which is determined according to a preset constraint condition and the to-be-planned intervention trajectory; the first motion state parameter model is used to represent a change relationship of a first motion state parameter of the vehicle from the current state to the target state with time; and planning a final intervention trajectory based on a Bezier curve according to the optimal parameter value and the first motion state parameter model. The present application is suitable for planning an intervention trajectory.

Description

Technical Field

[0001] This application relates to the field of vehicle-assisted driving technology, and in particular to a vehicle intervention trajectory planning method, device, electronic device, and readable storage medium. Background Technology

[0002] Autonomous driving, through communication between vehicles and between vehicles and infrastructure, can achieve path planning and safety precautions, reduce the burden on drivers, reduce traffic accidents caused by human factors, and make driving safer and easier.

[0003] Autonomous driving trajectory planning mainly refers to taking into account actual temporary or moving obstacles, speed, dynamic constraints, and driver comfort, and outputting an intervention trajectory as a function of time and position to enable intelligent driving vehicles to complete the predetermined driving task.

[0004] The current Intervention Trajectory Calculation (ITC) algorithm is based on Bézier curves and contains multiple constraints. Even so, it still includes multiple unknowns. In order to solve these unknowns, existing technologies require further assumptions for the ITC algorithm. However, the intervention trajectory obtained based on these assumptions has poor accuracy and insufficient adaptability to the environment, resulting in low vehicle obstacle avoidance efficiency. Summary of the Invention

[0005] In view of this, embodiments of this application provide a vehicle intervention trajectory planning method, apparatus, electronic device, and readable storage medium to improve the environmental adaptability of the intervention trajectory.

[0006] In a first aspect, embodiments of this application provide a vehicle intervention trajectory planning method, comprising: acquiring the current state and target state of the vehicle; determining preferred parameter values ​​for the parameters to be optimized based on a preset genetic algorithm, initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state; wherein the planned intervention trajectory is a Bézier curve-based intervention trajectory, and the parameters to be optimized are the parameters to be optimized for a first motion state parameter model in the planned intervention trajectory determined according to preset constraints and the planned intervention trajectory; the first motion state parameter model is used to represent the time-varying relationship of the first motion state parameters of the vehicle from the current state to the target state; and planning the final Bézier curve-based intervention trajectory based on the preferred parameter values ​​and the first motion state parameter model.

[0007] According to a specific implementation of this application, the step of determining the preferred parameter value of the parameter to be optimized based on a preset genetic algorithm, the initial value of the parameter to be optimized for the planned intervention trajectory, the current state, and the target state includes: determining multiple sets of initial values ​​for the parameter to be optimized for the planned intervention trajectory; determining the fitness value corresponding to each set of initial values ​​based on the multiple sets of initial values, the current state, and a preset fitness function; the fitness function is a function established based on the target state; incrementing the current iteration number by 1 to obtain a new iteration number; determining whether the new iteration number is greater than or equal to a preset threshold; if the new iteration number is not greater than and not equal to the preset threshold, generating multiple sets of candidate values ​​based on the multiple sets of initial values, using the multiple sets of candidate values ​​as new multiple sets of initial values, and iterating cyclically until the new iteration number is greater than or equal to the preset threshold; determining the candidate values ​​of the candidate value group with the smallest fitness value among the final multiple sets of candidate values ​​as the preferred value of the parameter to be optimized; wherein, the final multiple sets of candidate values ​​are the multiple sets of candidate values ​​when the new iteration number is greater than or equal to the preset threshold.

[0008] According to a specific implementation of this application, determining multiple initial values ​​of the parameters to be optimized for the planned intervention trajectory includes: determining multiple initial values ​​of the parameters to be optimized for the planned intervention trajectory based on a preset maximum first motion state parameter value and a preset maximum duration.

[0009] According to a specific implementation of an embodiment of this application, the plurality of initial values ​​includes a first initial value group; determining the fitness value corresponding to each initial value in the plurality of initial values ​​based on the plurality of initial values, the current state, and a preset fitness function includes: determining the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and the preset fitness function; wherein, determining the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and the preset fitness function includes: determining a change model of the second motion state parameters of the intervention trajectory to be planned based on the first motion state parameter model; determining the change amount of the second motion state parameter corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the change amount model of the second motion state parameters; and determining the fitness value corresponding to the first initial value group based on the change amount of the second motion state parameter and the preset fitness function.

[0010] According to a specific implementation of an embodiment of this application, determining the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and a preset fitness function further includes: determining the duration corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the duration sub-model of the first motion state parameter model; wherein, determining the fitness value corresponding to the first initial value group based on the change in the second motion state parameter and the preset fitness function includes: determining the fitness value corresponding to the first initial value group based on the change in the second motion state parameter, the duration, and the preset fitness function.

[0011] According to a specific implementation of an embodiment of this application, the first motion state parameter model includes multiple model segments; wherein, determining the change model of the second motion state parameter in the planned intervention trajectory based on the first motion state parameter model includes: determining the change model of the second motion state parameter corresponding to each model segment based on the multiple model segments of the first motion state parameter model; determining the change of the second motion state parameter corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter includes: determining the change of the second motion state parameter corresponding to each model segment based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter corresponding to each model segment; calculating the sum of the changes of the second motion state parameter corresponding to each model segment to obtain the total change of the second motion state parameter; and determining the total change of the second motion state parameter as the change of the second motion state parameter corresponding to the first initial value group.

[0012] According to a specific implementation of an embodiment of this application, the first motion state parameter is lateral acceleration; the second motion state parameter includes lateral velocity, lateral displacement and / or the rate of change of lateral acceleration.

[0013] According to a specific implementation of an embodiment of this application, the step of generating multiple candidate values ​​based on multiple initial values ​​includes: randomly selecting a preset number of initial value groups from the multiple initial value groups, and determining the initial value group with the largest fitness value among the preset number of initial value groups as a candidate value group; and / or, determining the candidate value group based on two initial values ​​from the multiple initial value groups and the weights corresponding to the two initial values; and / or, determining a target initial value group from the multiple initial value groups; and modifying the initial values ​​in the target initial value group to obtain the candidate value group.

[0014] Secondly, embodiments of this application provide a vehicle intervention trajectory planning device, comprising: an acquisition module for acquiring the current state and target state of the vehicle; a first determination module for determining preferred parameter values ​​of the parameters to be optimized based on a preset genetic algorithm, initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state; wherein the planned intervention trajectory is a Bézier curve-based planned intervention trajectory, and the parameters to be optimized are the parameters to be optimized of a first motion state parameter model in the planned intervention trajectory determined according to preset constraints and the planned intervention trajectory; the first motion state parameter model is used to represent the change relationship of the first motion state parameters of the vehicle from the current state to the target state over time; and a planning module for planning the final intervention trajectory based on the preferred parameter values ​​and the first motion state parameter model.

[0015] According to a specific implementation of an embodiment of this application, the first determining module includes: a first determining submodule, used to determine multiple initial values ​​of the parameters to be optimized for the planned intervention trajectory; a second determining submodule, used to determine the fitness value corresponding to each initial value in the multiple initial values ​​according to the multiple initial values, the current state, and a preset fitness function; the fitness function is a function established based on the target state; a counting submodule, used to increment the current iteration number by 1 to obtain a new iteration number; a judging submodule, used to judge whether the new iteration number is greater than or equal to a preset threshold; a loop iteration submodule, used to generate multiple candidate values ​​according to the multiple initial values ​​if the new iteration number is not greater than and not equal to the preset threshold, and use the multiple candidate values ​​as new multiple initial values, looping until the new iteration number is greater than or equal to the preset threshold; a third determining submodule, used to determine the candidate values ​​of the candidate value group with the smallest fitness value in the final multiple candidate values ​​as the preferred values ​​of the parameters to be optimized; wherein, the final multiple candidate values ​​are the multiple candidate values ​​when the new iteration number is greater than or equal to the preset threshold.

[0016] According to a specific implementation of an embodiment of this application, the first determining submodule is specifically used to: determine multiple initial values ​​of the parameters to be optimized for the planned intervention trajectory based on the preset maximum first motion state parameter value and the preset maximum duration.

[0017] According to a specific implementation of an embodiment of this application, the multiple sets of initial values ​​include a first set of initial values; the second determining submodule includes: a first determining unit, configured to determine the fitness value corresponding to the first set of initial values ​​based on the first set of initial values, the current state, and a preset fitness function; wherein, the first determining unit includes: a first determining subunit, configured to determine a change model of the second motion state parameters of the intervention trajectory to be planned based on the first motion state parameter model; a second determining subunit, configured to determine the change amount of the second motion state parameters corresponding to the first set of initial values ​​based on the first set of initial values, the current state, the target state, and the change model of the second motion state parameters; and a third determining subunit, configured to determine the fitness value corresponding to the first set of initial values ​​based on the change amount of the second motion state parameters and a preset fitness function.

[0018] According to a specific implementation of an embodiment of this application, the first determining unit further includes: a fourth determining subunit, configured to determine the duration corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the duration submodel of the first motion state parameter model; wherein, the third determining subunit is specifically configured to: determine the fitness value corresponding to the first initial value group based on the change amount of the second motion state parameter, the duration, and a preset fitness function.

[0019] According to a specific implementation of an embodiment of this application, the first motion state parameter model includes multiple model segments; wherein, the first determining unit is specifically used for: determining a change model of the second motion state parameter corresponding to each model segment based on the multiple model segments of the first motion state parameter model; the first determining unit is specifically used for: determining the parameter change of the second motion state corresponding to each model segment based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter corresponding to each model segment; calculating the sum of the changes of the second motion state parameters corresponding to each model segment to obtain the total change of the second motion state parameter; and determining the total change of the second motion state parameter as the change of the second motion state parameter corresponding to the first initial value group.

[0020] According to a specific implementation of an embodiment of this application, the first motion state parameter is lateral acceleration; the second motion state parameter includes lateral velocity, lateral displacement and / or the rate of change of lateral acceleration.

[0021] According to a specific implementation of an embodiment of this application, the loop iteration submodule is specifically used for: randomly selecting a preset number of initial value groups from multiple initial value groups, and determining the initial value group with the largest fitness value among the preset number of initial value groups as a candidate value group; and / or, determining the candidate value group based on two initial value groups among the multiple initial value groups and the weights corresponding to the two initial value groups respectively; and / or, determining a target initial value group from the multiple initial value groups; and modifying the initial values ​​in the target initial value group to obtain the candidate value group.

[0022] Thirdly, embodiments of this application provide an electronic device, which includes: a housing, a processor, a memory, a circuit board, and a power supply circuit, wherein the circuit board is disposed inside the space enclosed by the housing, and the processor and the memory are disposed on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the above-mentioned electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, for executing the vehicle intervention trajectory planning method described in any of the foregoing implementations.

[0023] Fourthly, embodiments of this application provide a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the vehicle intervention trajectory planning method described in any of the foregoing implementations.

[0024] The vehicle intervention trajectory planning method, device, electronic device, and readable storage medium of this embodiment determine the preferred parameter values ​​of the parameters to be optimized based on a preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state. The parameters to be optimized are the parameters to be optimized in the first motion state parameter model determined according to preset constraints and the intervention trajectory model based on Bézier curves. Based on the preferred parameter values ​​and the first motion state parameter model, the final intervention trajectory is planned. The mutation operation of the genetic algorithm can skip local optima and find an approximate optimal solution, i.e., the preferred parameter values. Based on this, the final intervention trajectory based on Bézier curves obtained has high environmental adaptability and improves the vehicle obstacle avoidance efficiency. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1A flowchart illustrating a vehicle intervention trajectory planning method provided in an embodiment of this application; Figure 2 The lateral acceleration curve of the ITC trajectory; Figure 3 A flowchart illustrating a vehicle intervention trajectory planning method provided in a specific embodiment of this application; Figure 4 This is a graph showing the relationship between fitness value and iteration number obtained through simulation in a specific embodiment of this application; Figure 5 This is a five-segment Bezier curve of displacement versus time obtained through simulation in a specific embodiment of this application; Figure 6 This is a five-segment Bezier curve of velocity versus time obtained through simulation in a specific embodiment of this application; Figure 7 This is a five-segment Bezier curve of acceleration versus time obtained through simulation in a specific embodiment of this application; Figure 8 This is a five-segment Bezier curve of the rate of change of acceleration versus time obtained through simulation in a specific embodiment of this application; Figure 9 This is a schematic diagram of the structure of a vehicle intervention trajectory planning device provided in an embodiment of this application; Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0027] The embodiments of this application will now be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are merely some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0028] Therefore, it is evident that in the existing technology, further assumptions are needed to obtain the final solution and the accuracy of the intervention trajectory obtained based on the assumptions.

[0029] To enable those skilled in the art to better understand the technical concept, implementation scheme and beneficial effects of the embodiments of this application, detailed descriptions are provided below through specific embodiments.

[0030] Figure 1 This is a flowchart illustrating a vehicle intervention trajectory planning method provided in an embodiment of this application, as shown below. Figure 1 As shown, the vehicle intervention trajectory planning method in this embodiment may include: S101. Obtain the current state and target state of the vehicle.

[0031] The current state may include the current position, current velocity, and / or current acceleration; the target state may include the target position, target velocity, and / or target acceleration.

[0032] In one embodiment, the current state may include the current lateral position, the current lateral velocity, and / or the current lateral acceleration; the target state may include the target lateral position, the target lateral velocity, and / or the target lateral acceleration.

[0033] S102. Based on the preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state, determine the optimal parameter values ​​for the parameters to be optimized.

[0034] This embodiment uses a genetic algorithm to determine the optimal parameter values ​​for the parameters to be optimized. This avoids the problem of inaccurate parameter values ​​caused by the assumptions used in existing technologies, which leads to poor environmental adaptability of the planned interference trajectory.

[0035] Genetic algorithms are heuristic optimization algorithms, such as particle swarm optimization and simulated annealing.

[0036] In this embodiment, the intervention trajectory to be planned is an intervention trajectory based on Bézier curves.

[0037] The parameters to be optimized are the parameters to be optimized in the first motion state parameter model of the planned intervention trajectory determined according to the preset constraints and the planned intervention trajectory; the first motion state parameter model is used to represent the relationship between the first motion state parameters of the vehicle from the current state to the target state and time.

[0038] The intervention trajectory model based on Bézier curves can be composed of multiple Bézier curve segments. Each segment includes four control points, and constraints are set to reduce algorithm complexity and adapt to practical needs. Specifically, the constraints can be that the first two control points of each acceleration curve segment are the same, the last two control points are the same, and the last control point of the previous Bézier curve segment is the same as the first control point of the next Bézier curve segment, etc.

[0039] Based on the preset constraints, the number of parameters to be solved in the first motion state parameter model can be reduced. In this embodiment, the parameters to be solved are taken as the parameters to be optimized.

[0040] The parameters to be optimized in this embodiment may include the first motion parameters at different times and / or the duration of each curve segment.

[0041] The first motion state parameter can be one of a variety of motion state parameters. Specifically, the motion state parameters can be velocity, acceleration, displacement, etc., and the first motion state parameter is one of them.

[0042] In one specific embodiment, the Intervention Trajectory Calculation (ITC) algorithm is obtained based on Bézier curves. Specifically, the vehicle state at each time point of the ITC curve can be represented by the following state variables: in, p represents the lateral deviation (vertical component) of the vehicle from the lane line. v is the lateral velocity (vertical component) of the vehicle relative to the lane line. a represents the lateral acceleration (vertical component) of the vehicle relative to the lane line; j represents the rate of change of acceleration of the vehicle relative to the lane line (jerk).

[0043] The entire trajectory is represented using piecewise curves, and the piecewise curves are defined as follows: as follows:

[0044] in, The total time taken for the trajectory to travel from the starting point to the ending point. equal , and Let K be the start and end times of the curve segment k.

[0045] when At that time, the curve needs to satisfy the following constraints: state variables It is a continuous function, that is , , , They are all continuous functions; Curve start point ; Curve endpoint ; in, : A function of the lateral deviation (vertical component) of the vehicle from the lane line as a function of time t; : A function of the lateral velocity (vertical component) of a vehicle relative to the lane line as a function of time t; : A function of the lateral acceleration (vertical component) of a vehicle relative to the lane line as a function of time t; Jerk is a function of the rate of change of acceleration of a vehicle relative to the lane line as a function of time t. To satisfy the definition of a Bézier curve, a coordinate system transformation is first required, which involves transforming the state vector... , , , Transform to a coordinate system based on variable s. , , , , where s is defined as follows:

[0046] When t is When the range changes, s takes values ​​from 0 to 1.0.

[0047] For ease of explanation, the following description uses Bezier curves to fit the acceleration information of each segment. Other variables include velocities such as displacement and the rate of change of acceleration, which is obtained by integrating or differentiating the acceleration.

[0048] Define the acceleration curve as a third-order Bézier curve:

[0049] in, d is a function of the lateral acceleration (vertical component) of the vehicle relative to the lane line as a function of time t; d is the Bessel order; k is the curve segment index. This is the i-th control point of the k-th segment of the Bézier curve; Bernstein basis polynomials are defined as follows:

[0050] in,

[0051] Integrating the lateral acceleration, the lateral velocity can be obtained. Horizontal distance lateral acceleration rate of change :

[0052]

[0053]

[0054] in, It is an arbitrary constant.

[0055] like Figure 2 As shown, the lateral acceleration of the ITC trajectory is composed of five segments of third-order lateral acceleration Bézier curves, with the duration of each Bézier curve being... , , , , .

[0056] According to the definition of Bézier curves, when s is 0, the starting point information of the curve is obtained; when s is 1, the ending point information is obtained. Therefore, the formula for the change in acceleration of segment k of the curve is as follows:

[0057]

[0058] Based on the change in acceleration of curve segment k, the formula for the change in velocity of curve segment k is as follows:

[0059] Similarly, the formula for calculating the displacement change of the k-segment curve is as follows:

[0060] The initial lateral acceleration is The lateral acceleration at the termination point is , The lateral acceleration during this period is , The lateral acceleration during this period is ,in for , for .

[0061] Each acceleration curve contains four control points and includes the following constraints: (1) The first two control points of each acceleration curve are the same, and the last two control points are the same, that is , ; (2) The last control point of the first segment of the Bézier curve is the same as the first control point of the second segment of the Bézier curve; (3) The maximum lateral acceleration change rate jerk shall not exceed the preset maximum lateral acceleration change rate. ; (4) The rate of change of lateral acceleration at the starting point of each acceleration curve segment and the rate of change of lateral acceleration at the termination point All are 0.

[0062] Since the Bézier curve is of the third order, the formula for the rate of change of acceleration of the k-segment curve is as follows:

[0063] Let its derivative be 0, and ,

[0064]

[0065] Substituting s=1 / 2 into... achievable

[0066]

[0067] This shows that if we can obtain , , , Then we can obtain , , Therefore, it is also necessary to solve for the unknowns. , , , Among them , , , These are the parameters to be optimized.

[0068] S103. Based on the preferred parameter values ​​and the first motion state parameter model, plan the final intervention trajectory based on Bézier curves.

[0069] In this embodiment, after obtaining the preferred parameter values, the relationship between the first motion state parameters and time can be obtained, and based on this, the final intervention trajectory can be planned.

[0070] In solving for the parameters to be optimized, the existing technology requires the following assumptions: (1) The duration of the second Bézier curve is equal to the duration of the fourth Bézier curve; (2) The lateral acceleration of the second Bézier curve is equal to the negative of the lateral acceleration of the fourth Bézier curve. Then, a larger lateral acceleration of the second Bézier curve is set according to the table. If the parameters to be optimized can be solved based on the lateral acceleration of the second Bézier curve, the current lateral acceleration of the second Bézier curve is used. Otherwise, it is reduced by a certain amount. Then, it is iterated again based on the reduced lateral acceleration of the second Bézier curve until a solution can be found. Otherwise, the planning fails.

[0071] In this embodiment, the optimal parameter values ​​of the parameters to be optimized are determined based on the genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state. The mutation operation of the genetic algorithm can skip local optima and find an approximate optimal solution. Based on this, the final intervention trajectory based on Bézier curves is highly adaptable to the environment and improves the vehicle obstacle avoidance efficiency.

[0072] In this embodiment, based on a preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state, the preferred parameter values ​​for the parameters to be optimized are determined. The parameters to be optimized are the parameters of the first motion state parameter model determined according to preset constraints and the intervention trajectory model based on Bézier curves. Based on the preferred parameter values ​​and the first motion state parameter model, the final intervention trajectory is planned. The mutation operation of the genetic algorithm can skip local optima and find an approximate optimal solution, i.e., the preferred parameter values. Based on this, the final intervention trajectory based on Bézier curves obtained has high environmental adaptability, improves vehicle obstacle avoidance efficiency, and avoids the problem of low accuracy of intervention trajectories caused by planning intervention trajectories based on assumptions in the prior art.

[0073] In an optional embodiment, determining the preferred parameter value of the parameter to be optimized based on a preset genetic algorithm, the initial value of the parameter to be optimized for the planned intervention trajectory, the current state, and the target state (S103) may include: S103a, Determine multiple initial values ​​for the parameters to be optimized for the planned intervention trajectory.

[0074] The number of parameters to be optimized can be one or more. If there are multiple parameters to be optimized, then each set of initial values ​​includes multiple initial values, that is, each parameter to be optimized corresponds to one initial value.

[0075] When each set of initial values ​​includes multiple initial values, each initial value in each set needs to be encoded. Specifically, the order of the initial values ​​in each set can be determined, and the order of the initial values ​​in each set should be the same.

[0076] S103b. Based on multiple initial values, the current state, and the preset fitness function, determine the fitness value corresponding to each initial value in the multiple initial values.

[0077] The fitness function is a function established based on the target state. The fitness function can be used to evaluate the quality of each solution (such as minimizing cost or maximizing profit). In this embodiment, the fitness function is established based on the target state.

[0078] S103c: Increment the current iteration count by 1 to get the new iteration count.

[0079] In this embodiment, the number of iterations is used as the termination condition for the loop iteration. Each time steps S103a and S103b are executed, the number of iterations is increased by 1 based on the current number of iterations to obtain a new number of iterations.

[0080] S103d: Determine whether the new iteration number is greater than or equal to the preset threshold.

[0081] S103e If the new iteration number is not greater than and not equal to the preset threshold, then generate multiple candidate values ​​based on multiple initial values, use the multiple candidate values ​​as new multiple initial values, and iterate cyclically until the new iteration number is greater than or equal to the preset threshold.

[0082] If the new iteration number is not greater than and not equal to the preset threshold, multiple sets of candidate values ​​are generated, and these multiple sets of candidate values ​​are used as new multiple sets of initial values. Then, the process jumps to step S103a to continue execution.

[0083] In embodiment A, generating multiple sets of candidate values ​​based on multiple sets of initial values ​​may include: Randomly select a preset number of initial value groups from multiple initial value groups, and determine the candidate value group by selecting the initial value group with the largest fitness value among the preset number of initial value groups.

[0084] For example, three sets of initial values ​​can be randomly selected, and the set with the largest fitness value among these three sets of initial values ​​can be determined as the candidate value set.

[0085] In embodiment B, generating multiple sets of candidate values ​​based on multiple sets of initial values ​​may include: Candidate value groups are determined based on two initial values ​​from multiple initial value groups and the weights corresponding to each initial value group.

[0086] For example, using a mixed crossover, each candidate value in the candidate value group is calculated as follows: each candidate value = alpha * first initial value + (1-alpha) * second initial value, where the first initial value is one of the two initial values, the second initial value is the other initial value, and alpha is the degree of mixing, which can take any value between 0 and 1, preferably alpha is 0.5.

[0087] In embodiment C, generating multiple sets of candidate values ​​based on multiple sets of initial values ​​may include: Determine the target initial value set from multiple initial value sets; Modify the initial values ​​in the target initial value group to obtain the candidate value group.

[0088] In this embodiment, if the target initial value group includes multiple initial values, some or all of the initial values ​​can be modified.

[0089] It is understood that two or three of the above embodiments A, B, and C can constitute a new embodiment for generating multiple sets of candidate values. In this case, if the goal is to generate m sets of candidate values, the number of candidate value sets generated by embodiments A, B, and C can be allocated according to a certain ratio.

[0090] S103f: Determine the preferred values ​​of the candidate value group with the smallest fitness value among the final multiple candidate values ​​as the preferred values ​​of the parameter to be optimized.

[0091] The final set of candidate values ​​is the set of candidate values ​​when the new iteration number is greater than or equal to the preset threshold.

[0092] In a specific example, determining multiple sets of initial values ​​(S103a) for the parameters to be optimized for the planned intervention trajectory may include: A1. Based on the preset maximum first motion state parameter value and the preset maximum duration, determine multiple sets of initial values ​​for the parameters to be optimized for the planned intervention trajectory.

[0093] For example, if the first motion state parameter is lateral acceleration, the maximum value of the first motion state parameter is... The maximum time it takes for the vehicle to travel from its current state to the target state, i.e., the maximum duration, is: The parameters to be optimized include lateral acceleration and duration. The initial value of lateral acceleration can be from 0 to... Choose from between these options; the initial value for the duration can be from 0 to... Furthermore, to adapt to actual driving environments, in some cases, the vehicle trajectory planning direction can be obtained by selecting from the available options. If the planned direction is to the left, then If to the right Furthermore, the lateral acceleration included in the parameters to be optimized comprises the lateral acceleration at two time points. Therefore, the initial value of a lateral acceleration can range from 0 to... Choose between these two options; the initial value of the other lateral acceleration can be from 0 to... Choose from the options.

[0094] For the calculation of fitness values ​​corresponding to a set of initial values, in some cases, multiple sets of initial values ​​may include the first set of initial values; Based on multiple sets of initial values, the current state, and a preset fitness function, the fitness value corresponding to each set of initial values ​​is determined (S103b), which may include: B1. Determine the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and the preset fitness function.

[0095] The process of determining the fitness value (B1) corresponding to the first initial value set based on the first initial value set, the current state, and the preset fitness function may include: B11. Based on the first motion state parameter model, determine the change model of the second motion state parameter in the planned intervention trajectory.

[0096] The change model of the second motion state parameters can be obtained by performing operations such as differentiation and integration on the first motion state parameter model.

[0097] The first motion state parameter is lateral acceleration; the second motion state parameter includes lateral velocity, lateral displacement, and / or the rate of change of lateral acceleration.

[0098] If the first motion state parameter model is a lateral acceleration model, and the second motion state parameter change model is a lateral velocity model, a lateral displacement model, and / or a lateral acceleration change rate model, then the lateral velocity change model can be obtained by integrating the lateral acceleration model; the lateral displacement change model can be obtained by integrating the lateral acceleration model twice; and the lateral acceleration change rate model can be obtained by differentiating the lateral acceleration model.

[0099] B12. Based on the first initial value group, the current state, the target state, and the change model of the second motion state parameters, determine the change of the second motion state parameters corresponding to the first initial value group.

[0100] The model for the change of the second motion state parameter is related to the first motion state parameter and time. Therefore, the change of the second motion state parameter corresponding to the first initial value set can be determined based on the first initial value set of the first motion state parameter, the current state, and the target state.

[0101] B13. Determine the fitness value corresponding to the first initial value group based on the change in the second motion state parameters and the preset fitness function.

[0102] It is understandable that the preset fitness function is related to the changes in the target state and the second motion state parameters.

[0103] To make the fitness values ​​more accurate, in some examples, determining the fitness value corresponding to the first initial value set based on the first initial value set, the current state, and a preset fitness function may further include: B14. Determine the duration corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the duration sub-model of the first motion state parameter model.

[0104] Since the duration sub-model is a model represented by the first motion state parameters at different times, the duration corresponding to the first initial value group can be determined after determining the first initial value group, the current state, and the target state.

[0105] The determination of the fitness value (B13) corresponding to the first initial value group based on the change in the second motion state parameter and the preset fitness function may include: B13a. Determine the fitness value corresponding to the first initial value group based on the change in the second motion state parameters, the duration, and the preset fitness function.

[0106] The preset fitness function is related to the target state, the change in the second motion state parameters, and the duration.

[0107] In cases where the first motion state parameter model comprises multiple model segments, in some examples, determining the second motion state parameter change model (B11) in the planned intervention trajectory based on the first motion state parameter model may include: B11a. Based on the multiple model segments of the first motion state parameter model, determine the change model of the second motion state parameter corresponding to each model segment.

[0108] In some examples, the first motion state parameter model is composed of five segments of third-order lateral acceleration Bezier curves pieced together.

[0109] Specifically, determining the change in the second motion state parameter corresponding to the first initial value set (B12) based on the first initial value set, the current state, the target state, and the change model of the second motion state parameter can include: B12a. Based on the first initial value group, the current state, the target state, and the change model of the second motion state parameters corresponding to each model segment, determine the change amount of the second motion state parameters corresponding to each model segment.

[0110] B12b. Calculate the sum of the changes in the second motion state parameters corresponding to each model segment to obtain the total changes in the second motion state parameters.

[0111] In this embodiment, the change in the second motion state parameter corresponding to each model segment is calculated, and the total change in the second motion state parameter is obtained by adding the changes in the second motion state parameters corresponding to each model segment.

[0112] B12c. Determine the total change in the second motion state parameter as the change in the second motion state parameter corresponding to the first initial value set.

[0113] The following detailed description of the solution in this application is based on a specific embodiment.

[0114] This embodiment is for the following... Figure 1 The example shown is an ITC trajectory whose lateral acceleration curve is composed of five segments of third-order lateral acceleration Bezier curves. Each segment contains four control points, with the x-axis representing time and the y-axis representing lateral acceleration.

[0115] When the intervention algorithm is composed of five segments of third-order lateral acceleration Bézier curves, the parameters to be optimized include: , , and .

[0116] The formula for the change in acceleration of the curve segment k (where k is an integer greater than or equal to 1 and less than or equal to 5) is as follows:

[0117]

[0118] The formula for the change in velocity of the k-segment curve is as follows:

[0119] The formula for the displacement change of the k-segment curve is as follows:

[0120] The formulas for the duration of each segment are as follows:

[0121] Among them This is the maximum value of the lateral acceleration rate of change, i.e., the threshold value of the lateral acceleration rate of change.

[0122] See Figure 3 The vehicle intervention trajectory planning method in this embodiment may include: Step 1: Start trajectory planning.

[0123] Typically, when there is an obstacle ahead, the driver needs to change lanes, or the vehicle deviates from its lane, the vehicle's intelligent driving system decision module issues an intervention command and then begins to plan the intervention trajectory.

[0124] Step 2: Obtain the current status of the vehicle, the planned trajectory direction of the vehicle, and the target state that the vehicle needs to achieve.

[0125] Establish a coordinate system with the center point of the rear axle of the vehicle as the origin. The positive X-axis is the forward direction of the vehicle, and the positive Y-axis is the leftward direction of the vehicle.

[0126] Get the lateral position of the vehicle at its current state. lateral velocity The lateral acceleration is the initial value. , Obtain the direction of vehicle trajectory planning If to the left If to the right .

[0127] Based on upstream perception and decision-making information, the target lateral position that the vehicle needs to reach is obtained. Target lateral velocity Target lateral acceleration .

[0128] Step 3: Set the number of individuals in the population m, the maximum number of iterations n, and initialize the number of iterations i=0.

[0129] Set the number of individuals in the genetic algorithm population to m, the maximum number of iterations to n, and the number of iterations to i.

[0130] Genetic Algorithm (GA) is an optimization algorithm inspired by biological evolution, used to solve complex search and optimization problems.

[0131] The formation of the population in a genetic algorithm: At the beginning, a set of possible solutions (called "individuals" or "chromosomes") is randomly generated to form a "population".

[0132] In this embodiment, the parameters to be optimized include , , and The parameters to be optimized are assigned values ​​to form m individuals, i.e. m sets of initial values.

[0133] Step 4: Randomly generate m individuals.

[0134] Based on the parameter values ​​of the parameters to be optimized according to the ITC algorithm, the resulting individuals are [ , , , ], randomly generate m individuals, among which, , , , , For maximum lateral acceleration, This represents the maximum duration.

[0135] Step 5: Calculate the value at the end of the curve.

[0136] Calculate the lateral displacement value at the end of the lateral displacement curve. lateral velocity value at the end of the lateral velocity curve and the total duration of the lateral acceleration curve .

[0137] m individuals [ , , , Substituting these values ​​into the formulas for the displacement change, velocity change, and duration of each segment in the aforementioned ITC algorithm, we can obtain the following: ; ; .

[0138] Step 6: Calculate the fitness function value.

[0139] The fitness function can evaluate the quality of each set of initial values.

[0140] The fitness function in this embodiment is: Fitness = 10 * End-point lateral displacement error + 20 * End-point lateral velocity error + 0.1 * Total duration, specifically:

[0141] Step 7: Determine whether the iteration number i is greater than or equal to the maximum iteration number n.

[0142] If not, proceed to step 8; if yes, proceed to step 9.

[0143] Step 8: Generate new individuals.

[0144] Genetic algorithms are used to generate new individuals through selection, crossover, and mutation operations until the number of newly generated individuals equals m. Then, the process jumps to step 5.

[0145] The selection operation involves choosing the better individual as the "parent" based on the fitness value.

[0146] Specifically, a tournament selection method is used: three individuals are randomly selected, and those with higher fitness are saved until a sufficient number are saved.

[0147] Crossover: mimics biological gene recombination, where two "parents" exchange parts of their code to generate a new individual (offspring).

[0148] Specifically, a mixed crossover is used, and the gene value of each offspring = alpha * parent + (1-alpha) * mother (alpha is the degree of mixing, which is generally 0.5).

[0149] Mutation: Randomly modifying certain coding bits of offspring to introduce diversity and prevent the algorithm from getting trapped in local optima.

[0150] Gaussian mutation is used to mutate each gene of the selected individuals, and then range constraints are applied to the mutated individuals.

[0151] Step 9: Output the trajectory corresponding to the optimal fitness value.

[0152] Encode the individual with the smallest fitness value among m individuals [ , , , ] as the preferred parameter values ​​for each parameter to be optimized, where for Figure 1 In , for Figure 1 In , Figure 1 In The lateral acceleration in the current state. For the lateral acceleration of the target state, after obtaining the optimal... , In the diagram and This can be determined, based on which, , and The calculation formula can be used to obtain , and The value of , thus, Figure 1 All values ​​of the curves in the curve are determined, that is, the relationship between lateral acceleration and time is determined. Based on this, the relationship between lateral displacement and time, the relationship between lateral velocity and time, and the relationship between the rate of change of lateral acceleration and time can be determined. Furthermore, based on the relationship between the four parameters of lateral acceleration, lateral displacement, lateral velocity, and rate of change of lateral acceleration and time, the disturbance trajectory can be planned.

[0153] In a specific example, the genetic algorithm parameters are set as follows: the number of individuals in the population is 100, the maximum number of iterations is 40, the selection probability is 0.1, the crossover probability is 0.7, and the mutation probability is 0.2.

[0154] Examples of simulation tests: The initial and target parameters of the ITC algorithm are shown in Table 1.

[0155] Table 1

[0156] The optimized fitness values ​​and the parameter values ​​to be optimized are shown in Table 2.

[0157] Table 2

[0158] Based on the above data, such as Figure 4-8 As shown, the simulation results include five-segment Bézier curves representing fitness value and iteration number, displacement and time, velocity and time, acceleration and time, and acceleration rate of change and time.

[0159] Figure 9This is a schematic diagram of the structure of a vehicle intervention trajectory planning device provided in an embodiment of this application, as shown below. Figure 9 As shown, the vehicle intervention trajectory planning device of this embodiment includes: an acquisition module 11, used to acquire the current state and target state of the vehicle; a first determination module 12, used to determine the preferred parameter value of the parameter to be optimized according to a preset genetic algorithm, the initial value of the parameter to be optimized of the intervention trajectory to be planned, the current state and the target state; wherein, the intervention trajectory to be planned is a Bézier curve-based intervention trajectory, and the parameter to be optimized is the parameter to be optimized of the first motion state parameter model in the intervention trajectory to be planned determined according to preset constraints and the intervention trajectory to be planned; the first motion state parameter model is used to represent the change relationship of the first motion state parameter of the vehicle from the current state to the target state over time; and a planning module 13, used to plan the final Bézier curve-based intervention trajectory according to the preferred parameter value and the first motion state parameter model.

[0160] The apparatus of this embodiment can be used to perform Figure 2 The technical solutions of the method embodiments shown are similar in principle and in effect, and will not be described again here.

[0161] The device in this embodiment determines the preferred parameter values ​​of the parameters to be optimized based on a preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state. The parameters to be optimized are the parameters of the first motion state parameter model determined based on preset constraints and the intervention trajectory model based on Bézier curves. Based on the preferred parameter values ​​and the first motion state parameter model, the device plans the final intervention trajectory. The genetic algorithm can obtain an approximate optimal solution for the parameters to be optimized, i.e., the preferred parameter values. The final intervention trajectory based on Bézier curves obtained has high environmental adaptability and improves the vehicle obstacle avoidance efficiency.

[0162] As an optional implementation, the first determining module includes: a first determining submodule, used to determine multiple sets of initial values ​​for the parameters to be optimized of the planned intervention trajectory; a second determining submodule, used to determine the fitness value corresponding to each set of initial values ​​in the multiple sets of initial values ​​according to the multiple sets of initial values, the current state, and a preset fitness function; the fitness function is a function established based on the target state; a counting submodule, used to increment the current iteration number by 1 to obtain a new iteration number; a judging submodule, used to judge whether the new iteration number is greater than or equal to a preset threshold; a loop iteration submodule, used to generate multiple sets of candidate values ​​according to the multiple sets of initial values ​​if the new iteration number is not greater than and not equal to the preset threshold, and use the multiple sets of candidate values ​​as new multiple sets of initial values, looping until the new iteration number is greater than or equal to the preset threshold; a third determining submodule, used to determine each candidate value of the candidate value group with the smallest fitness value in the final multiple sets of candidate values ​​as the preferred value of the parameters to be optimized; wherein, the final multiple sets of candidate values ​​are the multiple sets of candidate values ​​when the new iteration number is greater than or equal to the preset threshold.

[0163] As an optional implementation, the first determining submodule is specifically used to: determine multiple initial values ​​of the parameters to be optimized for the planned intervention trajectory based on the preset maximum first motion state parameter value and the preset maximum duration.

[0164] As an optional implementation, the multiple sets of initial values ​​include a first set of initial values; the second determining submodule includes: a first determining unit, configured to determine the fitness value corresponding to the first set of initial values ​​based on the first set of initial values, the current state, and a preset fitness function; wherein, the first determining unit includes: a first determining subunit, configured to determine a change model of the second motion state parameters of the intervention trajectory to be planned based on the first motion state parameter model; a second determining subunit, configured to determine the change amount of the second motion state parameters corresponding to the first set of initial values ​​based on the first set of initial values, the current state, the target state, and the change model of the second motion state parameters; and a third determining subunit, configured to determine the fitness value corresponding to the first set of initial values ​​based on the change amount of the second motion state parameters and a preset fitness function.

[0165] As an optional implementation, the first determining unit further includes: a fourth determining subunit, configured to determine the duration corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the duration submodel of the first motion state parameter model; wherein, the third determining subunit is specifically configured to: determine the fitness value corresponding to the first initial value group based on the change in the second motion state parameter, the duration, and a preset fitness function.

[0166] As an optional implementation, the first motion state parameter model includes multiple model segments; wherein, the first determining unit is specifically used to: determine the change amount model of the second motion state parameter corresponding to each model segment based on the multiple model segments of the first motion state parameter model; the first determining unit is specifically used to: determine the parameter change amount of the second motion state corresponding to each model segment based on the first initial value group, the current state, the target state, and the change amount model of the second motion state parameter corresponding to each model segment; calculate the sum of the change amounts of the second motion state parameters corresponding to each model segment to obtain the total change amount of the second motion state parameter; and determine the total change amount of the second motion state parameter as the change amount of the second motion state parameter corresponding to the first initial value group.

[0167] As an optional implementation, the first motion state parameter is lateral acceleration; the second motion state parameter includes lateral velocity, lateral displacement and / or the rate of change of lateral acceleration.

[0168] As an optional implementation, the iterative loop submodule is specifically used for: randomly selecting a preset number of initial value groups from multiple initial value groups, and determining the initial value group with the largest fitness value among the preset number of initial value groups as a candidate value group; and / or, determining the candidate value group based on two initial value groups among the multiple initial value groups and the weights corresponding to the two initial value groups respectively; and / or, determining a target initial value group from the multiple initial value groups; and modifying the initial values ​​in the target initial value group to obtain the candidate value group.

[0169] The apparatus described in the above embodiments can be used to execute the technical solutions of the above method embodiments. The implementation principle and technical effects are similar, and will not be repeated here.

[0170] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, as shown below. Figure 10 As shown, it may include a processor 10a and a memory 10b, wherein the memory 10b is used to store executable program code; the processor 10a runs the program corresponding to the executable program code by reading the executable program code stored in the memory 10b, and is used to execute the vehicle intervention trajectory planning method provided in the foregoing embodiment, so as to achieve the corresponding beneficial technical effects, which has been described in detail above and will not be repeated here.

[0171] The aforementioned electronic devices exist in various forms, including but not limited to: Accordingly, embodiments of this application also provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement any of the vehicle intervention trajectory planning methods provided in the foregoing embodiments, thus achieving the corresponding technical effects. This has been described in detail above and will not be repeated here.

[0172] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0173] The various embodiments in this specification are described in a related manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0174] In particular, the device embodiment is basically similar to the method embodiment, so the description is relatively simple. For relevant details, please refer to the description of the method embodiment.

[0175] For ease of description, the above apparatus is described by dividing it into various functional units / modules. Of course, in implementing this application, the functions of each unit / module can be implemented in one or more software and / or hardware.

[0176] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0177] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A vehicle intervention trajectory planning method, characterized in that, include: Obtain the current state and target state of the vehicle; Based on a preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state, the preferred parameter values ​​for the parameters to be optimized are determined; wherein, the planned intervention trajectory is a Bézier curve-based planned intervention trajectory, and the parameters to be optimized are the parameters to be optimized in the first motion state parameter model of the planned intervention trajectory determined according to preset constraints and the planned intervention trajectory; the first motion state parameter model is used to represent the change relationship of the first motion state parameters of the vehicle from the current state to the target state over time; Based on the preferred parameter values ​​and the first motion state parameter model, the final intervention trajectory based on Bézier curves is planned. The step of determining the preferred parameter value of the parameter to be optimized based on a preset genetic algorithm, the initial value of the parameter to be optimized for the planned intervention trajectory, the current state, and the target state includes: Determine multiple initial values ​​for the parameters to be optimized for the planned intervention trajectory; Based on multiple sets of initial values, the current state, and a preset fitness function, the fitness value corresponding to each set of initial values ​​is determined; the fitness function is a function established based on the target state. The multiple sets of initial values ​​include a first set of initial values; The step of determining the fitness value corresponding to each of the multiple initial values ​​based on the multiple initial values, the current state, and a preset fitness function includes: Based on the first initial value group, the current state, and the preset fitness function, determine the fitness value corresponding to the first initial value group; The step of determining the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and a preset fitness function includes: Based on the first motion state parameter model, determine the change model of the second motion state parameter in the planned intervention trajectory; Based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter, determine the change of the second motion state parameter corresponding to the first initial value group; Based on the first initial value group, the current state, the target state, and the duration sub-model of the first motion state parameter model, determine the duration corresponding to the first initial value group; The fitness value corresponding to the first initial value group is determined based on the change in the second motion state parameter, the duration, and the preset fitness function.

2. The method according to claim 1, characterized in that, After determining the fitness value corresponding to each of the multiple initial values, the step of determining the preferred parameter value of the parameter to be optimized based on the preset genetic algorithm, the initial value of the parameter to be optimized of the planned intervention trajectory, the current state, and the target state further includes: Increment the current iteration count by 1 to get the new iteration count; Determine whether the new iteration count is greater than or equal to a preset threshold; If the new iteration number is not greater than and not equal to the preset threshold, then based on the multiple initial values, multiple candidate values ​​are generated, and the multiple candidate values ​​are used as new multiple initial values. The iteration is repeated until the new iteration number is greater than or equal to the preset threshold. The candidate values ​​of the candidate value group with the smallest fitness value among the final multiple candidate values ​​are determined as the preferred values ​​of the parameter to be optimized; wherein, the final multiple candidate values ​​are the multiple candidate values ​​when the new iteration number is greater than or equal to the preset threshold.

3. The method according to claim 2, characterized in that, The multiple initial values ​​of the parameters to be optimized for determining the planned intervention trajectory include: Based on the preset maximum first motion state parameter value and the preset maximum duration, multiple sets of initial values ​​for the parameters to be optimized for the planned intervention trajectory are determined.

4. The method according to claim 1, characterized in that, The first motion state parameter model includes multiple model segments; The step of determining the change model of the second motion state parameter in the planned intervention trajectory based on the first motion state parameter model includes: Based on multiple model segments of the first motion state parameter model, determine the change model of the second motion state parameter corresponding to each model segment; The step of determining the change in the second motion state parameter corresponding to the first initial value group based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter includes: Based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter corresponding to each model segment, determine the change amount of the second motion state parameter corresponding to each model segment; Calculate the sum of the changes in the second motion state parameters corresponding to each model segment to obtain the total changes in the second motion state parameters; The total change in the second motion state parameter is determined as the change in the second motion state parameter corresponding to the first initial value group.

5. The method according to claim 1, characterized in that, The first motion state parameter is lateral acceleration; the second motion state parameter includes lateral velocity, lateral displacement and / or the rate of change of lateral acceleration.

6. The method according to claim 2, characterized in that, The step of generating multiple sets of candidate values ​​based on multiple sets of initial values ​​includes: Randomly select a preset number of initial value groups from multiple initial value groups, and determine the candidate value group by selecting the initial value group with the highest fitness value among the preset number of initial value groups; and / or, Candidate value groups are determined based on two initial values ​​from a set of multiple initial values ​​and their respective weights; and / or, Determine the target initial value set from multiple initial value sets; Modify the initial values ​​in the target initial value group to obtain the candidate value group.

7. A vehicle intervention trajectory planning device, characterized in that, include: The acquisition module is used to acquire the current state and target state of the vehicle. The first determining module is used to determine the preferred parameter values ​​of the parameters to be optimized based on a preset genetic algorithm, the initial values ​​of the parameters to be optimized for the planned intervention trajectory, the current state, and the target state; wherein, the planned intervention trajectory is a planned intervention trajectory based on a Bézier curve, and the parameters to be optimized are the parameters to be optimized in the first motion state parameter model of the planned intervention trajectory determined according to preset constraints and the planned intervention trajectory; the first motion state parameter model is used to represent the change relationship of the first motion state parameters of the vehicle from the current state to the target state over time; The planning module is used to plan the final intervention trajectory based on Bézier curves according to the preferred parameter values ​​and the first motion state parameter model. The first determining module includes: The first determining submodule is used to determine multiple sets of initial values ​​for the parameters to be optimized of the intervention trajectory to be planned; The second determining submodule is used to determine the fitness value corresponding to each of the multiple initial values ​​based on the multiple initial values, the current state, and a preset fitness function; the fitness function is a function established based on the target state. The multiple sets of initial values ​​include a first set of initial values; The second determining submodule is specifically used for: Based on the first initial value group, the current state, and the preset fitness function, determine the fitness value corresponding to the first initial value group; The step of determining the fitness value corresponding to the first initial value group based on the first initial value group, the current state, and a preset fitness function includes: Based on the first motion state parameter model, determine the change model of the second motion state parameter in the planned intervention trajectory; Based on the first initial value group, the current state, the target state, and the change model of the second motion state parameter, determine the change of the second motion state parameter corresponding to the first initial value group; Based on the first initial value group, the current state, the target state, and the duration sub-model of the first motion state parameter model, determine the duration corresponding to the first initial value group; The fitness value corresponding to the first initial value group is determined based on the change in the second motion state parameter, the duration, and the preset fitness function.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, wherein the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, for executing the vehicle intervention trajectory planning method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the vehicle intervention trajectory planning method according to any one of claims 1-6.

Images