An intelligent networked vehicle lane-changing trajectory planning method considering passenger individual ride demand

Abstract

The application relates to the technical field of automatic driving and discloses an intelligent networked vehicle lane-changing trajectory planning method considering passenger individualized riding demands. The method mainly comprises three parts, namely a passenger individualized demand extraction layer, a decision layer and a planning layer. In the passenger individualized demand extraction layer, a passenger demand factor mapping model based on an analytic hierarchy process and a fuzzy logic system is introduced, which is used for subsequent individualized quantitative benchmarks. In the decision layer, a lane-changing decision model considering passenger individualized riding demands is designed based on an artificial potential field theory and a pure strategy game method, which can generate individualized lane-changing decisions based on driving environment information and passenger demand factors. In the planning layer, a lane-changing duration prediction model is designed for lane-changing duration constraints, and the problem is converted into a quadratic programming problem, so that an individualized lane-changing trajectory meeting passenger demands is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a smart connected vehicle lane-changing trajectory planning method considering passenger individualized ride demand. BACKGROUND

[0002] Currently, the rapid development of automatic driving technology, especially the breakthrough of lane-changing trajectory planning technology, has become the core of smart connected vehicles integrating into complex traffic environments. In driving behavior, lane-changing trajectory planning not only concerns the safety of the vehicle itself, but also directly affects the overall traffic flow. Therefore, it is necessary to accurately analyze the operating state of surrounding vehicles to ensure that the lane-changing process is both fast and safe. The demand of passengers often directly affects the driving direction of the autonomous vehicle, and the automatic driving trajectory planning combined with passenger demand can lead the traffic system to a higher level of intelligentization. At present, there are many lane-changing trajectory planning methods for autonomous vehicles, such as optimization-based and learning-based methods. The optimization-based method essentially abstracts one or more objectives and designs corresponding constraints to achieve the purpose of trajectory planning; the learning-based method generally models the trajectory planning problem as a Markov process, and realizes the training of intelligent agents for different tasks by designing the interaction mode between intelligent agents and the environment.

[0003] Although existing methods have achieved certain results in ordinary lane-changing trajectory planning tasks, the existing methods do not fully study the passenger's ride preference characteristics, and few studies have considered the passenger's ride preference in the consideration of autonomous vehicle lane-changing trajectory planning. In addition, the current research on passenger driving preferences combined with lane-changing trajectory planning also has shortcomings. On the one hand, the existing method relies on learning from the environment and needs a large amount of passenger preference data; on the other hand, the existing method considers the lane-changing trajectory from the perspective of the driver, without considering the real needs of the passengers, so how to conveniently and quickly represent the specific needs of the passengers becomes a technical difficulty, in addition, how to deeply integrate the real needs of the passengers in the whole lane-changing decision and trajectory planning process is also a technical difficulty at present. SUMMARY

[0004] In order to solve the above technical problems, the present application provides a smart connected vehicle lane-changing trajectory planning method considering passenger individualized ride demand;

[0005] To achieve the above technical scheme, the following is specific:

[0006] S1, converting the ride preference degree of the passenger into a judgment matrix, and mapping the judgment matrix into a passenger preference weight vector;

[0007] The ride preference degree of the passenger includes: ride comfort preference degree, safety preference degree and efficiency preference degree;

[0008] The ride comfort preference degree, the safety preference degree and the efficiency preference degree are combined into a judgment matrix A in a 3x3 matrix form; wherein the element A ij represents the interaction result between the ride comfort, the safety and the efficiency according to the passenger;

[0009] The passenger preference weight vector P=[w c w s w e ] respectively represents the importance value of the passenger after the interaction of the comfort, the safety and the efficiency; the solving method is the analytic hierarchy process, and the specific steps include:

[0010] S1.1, the sum of each column of the judgment matrix is calculated; the expression is as follows:

[0011]

[0012] In the formula, S j represents the sum of the elements in the jth column of the judgment matrix A; A ij represents the element in the ith row and the jth column of the judgment matrix A, represents the sum of the elements in the jth column;

[0013] S1.2, each element in the judgment matrix A is standardized; the expression is as follows:

[0014]

[0015] In the formula, A' ij represents the standardized judgment matrix element;

[0016] S1.3, each weight value in the passenger preference weight vector is calculated; the expression is as follows:

[0017]

[0018] In the formula, P represents the passenger preference weight vector; P k represents the kth element in the passenger preference weight vector;

[0019] S1.4, the consistency of the passenger preference weight vector is checked to ensure the rationality of the judgment matrix; the specific steps are as follows:

[0020] S1.4.1, the maximum eigenvalue is calculated; the expression is as follows:

[0021]

[0022] In the formula, λ max is the maximum eigenvalue of the judgment matrix A, (A·P)k represents the kth element of the product of the judgment matrix A and the passenger preference weight vector P, and represents the comprehensive score of each factor;

[0023] S1.4.2, calculate the consistency index; the expression is as follows:

[0024]

[0025] In the formula, CI represents the consistency index value, and n represents the matrix dimension;

[0026] S2, establish a secondary mapping method to obtain the passenger preference factor f through the passenger preference weight vector;

[0027] The secondary mapping method includes a triangular membership function and a Gaussian membership function;

[0028] The specific process is represented as follows:

[0029] S2.1, define the triangular membership function:

[0030]

[0031] In the formula, a represents the left end point of the triangle in the triangular membership function, b represents the right end point of the triangle in the triangular membership function, c represents the peak point of the triangle in the triangular membership function, x represents the input of the triangular membership function, i.e. each weight value in the passenger preference weight vector P, and μ T (x) represents the triangular membership function with input x;

[0032] S2.2, define the Gaussian membership function for defining the fuzziness:

[0033]

[0034] In the formula, μ and σ respectively represent the mean and standard deviation of the Gaussian distribution, y represents the output value of the triangular membership function, and μ G (y) represents the Gaussian membership function with input y;

[0035] S2.3, input each weight value in the passenger preference weight vector P to the triangular membership function and then pass through the Gaussian membership function, so that the final passenger preference factor f can be calculated by summation; the expression is as follows:

[0036]

[0037] In the formula, the passenger preference factor f ∈ [-1, 1], which directly reflects the type of the passenger, including aggressive type, ordinary type and conservative type, corresponding to the passenger preference factor f = -1, f = 0 and f = 1 respectively;

[0038] The above steps achieve a self-supervised mapping mode of passenger preference weight factors, overcoming the technical difficulty of requiring a large amount of calibration data for supervised learning such as deep learning;

[0039] S3, through the three-layer personalized lane-changing decision-making model, a mapping relationship between the passenger acceptable collision time and the passenger demand factor is established, and a lane-changing trajectory planning decision of the ego vehicle is generated;

[0040] The three-layer personalized lane-changing decision-making model includes a risk assessment layer, a global benefit assessment layer, and a personalized assessment layer.

[0041] The establishment process is as follows:

[0042] S3.1, in the risk assessment layer, the artificial potential field theory is introduced, and the risk value of each lane is calculated, and the calculation formula is as follows:

[0043]

[0044] In the formula, represents the potential field size of the oth lane; r represents the number of all obstacle vehicles on the oth lane; t e represents the lane-changing duration; η obs is the lane weight; A μ is the longitudinal cross-sectional area of the μth obstacle vehicle, which is used to represent the size of the vehicle, μ∈(1, +∞); is the Euclidean distance between the ego vehicle and the μth obstacle vehicle at time t; Q represents the potential field influence range of the vehicle, that is, when the Euclidean distance between the ego vehicle and the obstacle vehicle exceeds Q, the potential of the vehicle in the current lane is not calculated. represents the relative speed of the ego vehicle and the μth obstacle vehicle at time t, due to the limitation of the planning frequency of 10Hz, the interval between time t and t+1 is 0.1s; in addition, o and r are integers in the interval range; the above process is only when the gap threshold ξ is greater than the gap threshold ξ, the ego vehicle generates a lane-changing intention;

[0045] The gap threshold ξ calculation process is as follows:

[0046] The minimum braking distance of the ego vehicle and the obstacle vehicle without collision is obtained; the expression is as follows:

[0047]

[0048] In the formula, a bmax represents the maximum braking deceleration of the ego vehicle, s represents the minimum braking distance of the ego vehicle and the obstacle vehicle without collision, at this time, the following can be obtained:

[0049]

[0050] That is, the potential energy field size of the oth lane can be regarded as a function of the minimum braking distance s, where s∈(0,+∞). The function is a convex function in the domain of definition. Considering safety, the present invention calculates the value of the minimum braking distance s that can make the potential energy field of the oth lane Minimum, that is, solving the following equation:

[0051]

[0052] Where, represents the differential of the potential energy field of lane o with respect to the minimum braking distance s for the ego vehicle to avoid collision with the obstacle vehicle. The solution of this equation is regarded as the gap threshold ξ;

[0053] S3.2. In the global evaluation layer, a pure strategy game method is used to generate an optimal strategy between the ego vehicle and the vehicle behind it in the target lane, which can minimize the impact on the overall traffic situation. The steps to obtain the optimal strategy are as follows:

[0054] S3.2.1. Generate a set of strategies with the highest efficiency between the ego vehicle and the vehicle behind it in the target lane using a pure strategy game method, and make a lane change decision.

[0055]

[0056] Where U VC (t) and U VT (t) represents the benefit value of the current lane and the target lane of the ego vehicle at time t, respectively; U(t) is the total benefit of the local traffic system at time t; Γ represents the number of all obstacle vehicles in front of the current lane of the ego vehicle; p and q are the number of all obstacle vehicles in front of and behind the target lane from the perspective of the ego vehicle, respectively; ω t is the time weight value at time t; V(t), and represents the vehicle, the γth obstacle vehicle in front of the current lane, the ρth obstacle vehicle in front of the target lane, and the ρth obstacle vehicle behind the target lane. The speed of the obstacle car at the t moment; where Γ∈[0,+∞), p∈[0,+∞), q∈[0,+∞), γ∈[0,+∞), ρ∈[0,+∞), The specific value depends on the perception limit of the vehicle. Therefore, the value of Γ, p and q is a dynamic process.

[0057] All strategy combinations include: "the ego vehicle changes lanes and the target lane rear vehicle accepts" and "the ego vehicle does not change lanes and the target lane rear vehicle accepts"; the global benefit value of each strategy combination is calculated based on the above global benefit function, and the evaluation is completed by finding the optimal benefit combination, it should be noted that the global benefit evaluation continues only when the ego vehicle strategy is lane changing, otherwise the decision of not changing lanes is directly output;

[0058] S3.3、If both layers make lane-changing strategies, further evaluate the passenger's demand by comparing the collision time between the ego vehicle and the target lane rear vehicle to determine whether the passenger's individualized demand is met;

[0059] Specifically, in the individualized evaluation layer, a mapping relationship between the passenger's minimum acceptable collision time and the passenger demand factor f is established, and the establishment process is as follows:

[0060] S3.3.1, predict the predicted collision time; the expression of the predicted collision time is as follows:

[0061]

[0062] In the formula, T A represents the predicted collision time; D TF is the Euclidean distance between the ego vehicle and the target lane rear vehicle; |v-v TF | represents the absolute difference in speed between the ego vehicle and the target lane rear vehicle; v represents the speed of the ego vehicle; v TF is the speed of the target lane rear vehicle;

[0063] S3.3.2, establish a mapping relationship between the passenger's acceptable collision time and the passenger demand factor f, and the expression is as follows:

[0064] In the formula, T P represents the passenger's acceptable collision time; T S is a time constant; η T represents the scaling weight, which is valued at 0.3 according to experience;

[0065] S3.3.3, get the final output through the predicted collision time and the mapping relationship;

[0066] Specifically, based on the above analysis, when T A >T P , it indicates that the predicted collision time with the target lane rear vehicle meets the passenger's demand, i.e. the final output allows the execution of the lane-changing decision; otherwise, the passenger rejects the lane-changing decision, and the lane-changing decision is not executed;

[0067] S4, the lane-changing trajectory planning decision problem of the ego vehicle is converted into solving a multivariate linear equation set of the lane-changing trajectory by establishing a trajectory generation model based on a multi-dimensional coordinate space; the expression is as follows:

[0068]

[0069] wherein, and respectively represent the projected distance on the target lane road center line, the initial speed, the initial acceleration, the initial planar transverse coordinate, the initial planar longitudinal coordinate and the slope of the path point at the initial time t s of the lane-changing of the ego vehicle; and respectively represent the projected distance on the target lane road center line, the speed at the end time, the acceleration at the end time, the planar transverse coordinate at the end time, the planar longitudinal coordinate at the end time and the slope of the path point at the end time t e of the lane-changing of the ego vehicle; in order to facilitate calculation, the above state quantities are combined into an initial state vector and an end state vector When the above two state vectors are determined, the trajectory polynomial coefficient vectors α and β can also be determined in the form of solving a multivariate linear equation set. Since all the state quantities in the initial state vector can be obtained by the vehicle-mounted sensor, it is necessary to further calculate the state quantities in the end state vector. As it is noted that the target lane center line is actually a sequence of discrete points, only the nearest point to needs to be searched by the Euclidean distance, and all the state quantities except the end time t e , the speed and the acceleration at the end time can be solved by linear interpolation between the previous point and the next point thereof;

[0070] S5, the unknown quantities of the lane-changing trajectory multivariate linear equation set are solved by optimization;

[0071] that is, by setting a cost function and constraint conditions considering passenger preferences, the unknown quantity solving problem of the lane-changing trajectory multivariate linear equation set is converted into a nonlinear optimization problem; the specific steps are as follows:

[0072] S5.1, a multi-objective cost function considering safety, comfort and efficiency is established according to passenger preferences, including the following steps:

[0073] S5.1.1, a lane-level potential field is introduced as a safety constraint, which comprehensively considers the relative speed, acceleration and longitudinal cross-sectional area of the vehicle along the y direction;

[0074] The lane-level potential field calculation expression is as follows:

[0075]

[0076] wherein, and respectively represent the projection distance value of the ego vehicle, the obstacle vehicle behind the ego vehicle in the target lane, the obstacle vehicle in front of the ego vehicle in the target lane and the obstacle vehicle in front of the ego vehicle in the current lane on the target lane at the t-th moment; and respectively represent the speed of the ego vehicle and the obstacle vehicle behind the ego vehicle in the target lane at the t-th moment; and are the maximum braking deceleration of the ego vehicle and the obstacle vehicle behind the ego vehicle in the target lane; τ is the planning time step of the ego vehicle; Q is the potential energy influence range of each vehicle;

[0077] S5.1.2, the safety cost function is satisfied by the second-order Taylor expansion at each trajectory point in the quadratic form; the expression is as follows:

[0078]

[0079]

[0080] wherein, Js represents the safety cost function; X represents the high-dimensional unknown form; X0 is the input of each trajectory point; is the gradient calculation; H represents the Hessian matrix; represents the partial differential calculation;

[0081] S5.1.3, the acceleration and jerk (acceleration change rate) of each trajectory point of the ego vehicle are introduced to represent on the comfort and efficiency cost function; the expression is as follows:

[0082]

[0083] wherein, J c and J e respectively represent the comfort cost function and the efficiency cost function; j(t) represents the jerk of the ego vehicle at the t-th moment; t e represents the total time length of the lane change;

[0084] S5.1.4, the multi-objective cost function is obtained in combination with the safety, comfort and efficiency; the expression is as follows:

[0085] J = w s · J s + w c · J c + w e · J e

[0086] where w s , w c , and w e are three index weight values in the passenger preference weight vector P;

[0087] S5.2, optimizing the hyperparameters of the long short-term memory neural network by the whale optimization algorithm to obtain the upper and lower limits of the safety constraint interval;

[0088] The specific steps are as follows:

[0089] S5.2.1, establishing a long short-term memory neural network based on the whale optimization algorithm optimization to obtain the estimated driving distance; the expression is as follows:

[0090]

[0091] where v is the speed of the ego vehicle at the initial time of lane changing, is the estimated driving distance, and v represent the speeds of the ego vehicle and the target lane in front of the obstacle vehicle, respectively;

[0092] S5.2.2, collecting a data set from an open source data set, wherein 70% is randomly selected as a training set and the remaining 30% is selected as a test set; and the hyperparameters of the long short-term memory neural network are optimized by the whale optimization algorithm, including the learning rate and the maximum number of iterations;

[0093] The loss value is calculated by the loss function, and the whale optimization algorithm needs to search for the value of the hyperparameter when the loss value is minimized to update the model weight in reverse:

[0094]

[0095] where Λ represents the total number of training samples; T φ and t represent the true lane changing time and the model predicted lane changing time, respectively; F represents the prediction loss value of each iteration, i.e. the fitness function value of the whale optimization algorithm;

[0096] S5.2.3, introducing a passenger riding preference factor f and designing the following lane changing time constraint form:

[0097] t e ∈ [T min , T max ]

[0098]

[0099] where T min represents the lower bound of the lane changing time, and T maxrepresents the upper bound of the lane change duration, Indicates the predicted lane change duration output by the model;

[0100] S5.2.4. Introduce collision constraints during vehicle movement to ensure that the vehicle does not collide;

[0101] Specifically, it is necessary to ensure that the vehicle does not collide during driving, so it is necessary to introduce collision constraints into the optimization problem; in order to improve the computational efficiency of the model, the collision constraints of the present invention can be expressed in a linearized form;

[0102] The distance projected on the target lane by the midpoint of the rear axle of the ego vehicle at the initial moment of lane change is The three vertices, namely the left side of the rear bumper, the left side of the front bumper, and the right side of the front bumper, can be calculated through geometric calculation as follows:

[0103]

[0104] Where w ego is the width of the vehicle; θ t is the heading angle of the vehicle at time t, θ t ∈(0,2π);lr ego Indicates the length from the midpoint of the vehicle's rear axle to the vehicle's rear bumper; and They represent the projection distance values ​​of the left point of the rear bumper, the left point of the front bumper, and the right point of the front bumper on the center line of the target lane at time t, respectively. and The value range of is [0,+∞); and Combined into a vector S Ω , then obviously S Ω is the projection value of the vehicle's distance on the center line of the target lane A linear function of χ∈{A,B,C}; in this case, the linearization constraint of the ego-vehicle safety can be expressed as:

[0105]

[0106] Where, f colli and r colli They represent the safety distance threshold in front of the vehicle and the safety distance threshold behind the vehicle respectively; s 1 ,t 、s 2,t and s 3,t They represent the projection value of the obstacle vehicle behind the target lane on the target lane, the projection value of the obstacle vehicle in front of the target lane on the target lane, and the projection value of the obstacle vehicle in front of the lane where the vehicle is currently located on the target lane;

[0107] S5.2.5, the planned trajectory needs to apply motion state constraints at the end of the lane change;

[0108] Specifically, in order to ensure that the ego vehicle can smoothly follow the traffic flow of the target lane at the end of the lane change, it is necessary to apply motion state constraints to the end point of the lane change trajectory planned by the ego vehicle;

[0109] The present application introduces a vehicle following model to calculate the motion state of the end point of the lane change, and derives the acceleration constraint at the end point of the lane change as follows:

[0110]

[0111] In the formula, denotes the vector composed of the accelerations of the obstacle vehicles behind and in front of the ego vehicle on the target lane; is the speed of the obstacle vehicle in front of the ego vehicle in the current lane at the end point of the lane change; Δs * is a distance constant, and the value range is Δs * ∈[1,5]; ω1 and ω2 are respectively the distance weight value and the relative speed weight value, the distance weight value value range is ω1∈[0,1]; the relative speed weight value value range is ω2∈[0,1]; l OV-3 and l ego respectively denote the length of the obstacle vehicle in front of the ego vehicle in the current lane and the length of the ego vehicle; therefore, the motion state constraint can be summarized as:

[0112]

[0113] In the formula, denotes the projection distance of the ego vehicle on the center line of the target lane at the t-th moment; denotes the speed of the ego vehicle at the t-th moment;

[0114] S5.3, the trajectory optimization problem described in the above steps S5.1 and S5.2 is arranged into a nonlinear programming problem, and the complete expression is as follows:

[0115]

[0116] The expression of the above nonlinear programming problem is brought into the IPOPT sequence quadratic programming solver to solve, and the personalized planning trajectory is obtained;

[0117] Advantages of the present application

[0118] The present application can be mainly divided into three parts, which are passenger personalized demand extraction layer, decision layer and planning layer. In the passenger personalized demand extraction layer, the present application introduces a passenger demand factor mapping model based on analytic hierarchy process and fuzzy logic system, which can map the language information at the passenger end into passenger demand weight vector and passenger demand factor, which is used for subsequent personalized quantitative benchmark. In the decision layer, the present application designs a lane changing decision model considering passenger personalized riding demand based on artificial potential field theory and pure strategy game method, which can generate personalized lane changing decision based on driving environment information and passenger demand factor. In the planning layer, the present application constructs a weighted cost function through passenger demand weight vector and converts the lane changing trajectory planning problem into a nonlinear optimization problem. In order to further improve the calculation efficiency, the present application designs a constraint linearization method for safety constraint, and introduces a long short-term memory neural network model optimized by optimization algorithm, which is used for the prediction of the lane changing time of the ego vehicle. The prediction result will be used as the prior condition of the trajectory planning. Finally, the personalized lane changing trajectory meeting the passenger demand is planned based on the quadratic programming solution of the lane changing trajectory planning problem, so as to improve the problem that the passenger personalized demand of the autonomous vehicle is difficult to meet, and optimize the operation of the overall traffic flow.

[0119] Compared with the prior art, first, the present application does not depend on calibration data which needs large cost, can realize the mapping logic of passenger fuzzy demand to fixed range preference weight value, can directly represent the specific demand of passengers conveniently and quickly, and has superior performance in calculation complexity compared with the prior art; second, the decision layer of the present application greatly reduces the calculation complexity compared with the prior art based on learning or optimization, and has great potential for online application; in addition, compared with the prior art, the present application considers the influence of the lane changing behavior of the ego vehicle on the global traffic running state, and optimizes the operation of the overall traffic flow to a certain extent; further, the present application models the personalized lane changing trajectory planning problem of the passenger as a sequential quadratic programming problem, and designs a linearization method of complex constraints, which has higher calculation efficiency compared with various trajectory planning methods based on optimization; finally, the real demand of the passenger is reflected in the lane changing decision and planning process of the ego vehicle through the passenger preference weight vector and the passenger preference factor, which can more deeply and systematically reflect the real demand of the passenger compared with the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0120] Figure 1 is the flowchart of the present application;

[0121] Figure 2 is the schematic diagram of the present application;

[0122] Figure 3 is the schematic diagram of the passenger preference extraction model of the present application;

[0123] Figure 4 The process of mapping the passenger preference weight vector to the passenger preference factor of the present invention;

[0124] Figure 5 Schematic diagram of the potential energy field range of the present invention;

[0125] Figure 6 This is a schematic diagram of the safety constraint linearization of the present invention;

[0126] Figure 7 This is a schematic diagram of the lane change duration prediction effect of the present invention;

[0127] Figure 8 Schematic diagram of the lane-changing trajectory planning results of the present invention in different test scenarios, where part (a) is a schematic diagram of the motion planning curves for three different passengers; part (b) is a schematic diagram of the lane-changing paths corresponding to different speeds and accelerations for the three different passengers. DETAILED DESCRIPTION

[0128] The present invention will be further described in detail below with reference to specific embodiments.

[0129] like Figure 1 , Figure 2 As shown, a lane-changing trajectory planning method for an intelligent connected vehicle considering the personalized riding needs of passengers includes the following steps:

[0130] S1. Convert the passenger's riding preference into a judgment matrix, and map the judgment matrix into a passenger preference weight vector;

[0131] like Figure 3 As shown, the passenger preference extraction model is used to extract the passenger's riding preference degree;

[0132] Passengers' riding preferences include: riding comfort preference, safety preference and efficiency preference;

[0133] The preference degree of riding comfort, safety and efficiency is formed into a judgment matrix A in the form of a 3×3 matrix; wherein the element A in the i-th (i=1, 2, 3) row and the j-th (j=1, 2, 3) column is ij It represents the interaction between passengers’ perception of ride comfort, safety and efficiency;

[0134] Passenger preference weight vector P = [w c w s w e ], which respectively represent the importance of the passengers to the interaction of comfort, safety and efficiency; the solution method is the hierarchical analysis method, and the specific steps include:

[0135] S1.1. Calculate the sum of each column of the judgment matrix; the expression is as follows:

[0136]

[0137] where S j denotes the sum of the elements in the jth column of the judgment matrix A; A ij denotes the element in the ith row and jth column of the judgment matrix A, denotes the sum of the elements in the jth column;

[0138] S1.2, standardize each element in the judgment matrix A; the expression is as follows:

[0139]

[0140] where A' ij denotes the element of the standardized judgment matrix;

[0141] S1.3, calculate each weight value in the passenger preference weight vector; the expression is as follows:

[0142]

[0143] where P denotes the passenger preference weight vector; P k denotes the kth element in the passenger preference weight vector;

[0144] S1.4, ensure the rationality of the judgment matrix by performing a consistency check on the passenger preference weight vector; the specific steps are as follows:

[0145] S1.4.1, calculate the maximum eigenvalue; the expression is as follows:

[0146]

[0147] where λ max is the maximum eigenvalue of the judgment matrix A, (A·P) k denotes the kth element of the product of the judgment matrix A and the passenger preference weight vector P, indicating the comprehensive score of each factor;

[0148] S1.4.2, calculate the consistency index; the expression is as follows:

[0149]

[0150] where CI denotes the consistency index value, and n denotes the matrix dimension;

[0151] If the consistency ratio CR < 0.1, it is considered that the passenger preference weight vector calculated in steps S1.1 to S1.3 is acceptable, otherwise it indicates that the judgment matrix A has a high consistency, i.e. the information input by the passenger has a large conflict, which is caused by input error, so the passenger preference weight vector obtained is not reasonable, at this time the passenger needs to re-input the ride preference and repeat steps S1.1 to S1.6 until the consistency ratio CR < 0.1;

[0152] S2, a secondary mapping method is established to obtain the passenger preference factor f through the passenger preference weight vector;

[0153] The process of mapping the passenger preference weight vector to the passenger preference factor is shown in Figure 4 , and is as follows:

[0154] The secondary mapping method includes a triangular membership function and a Gaussian membership function; that is, by establishing a fuzzy logic system coupled by the triangular membership function and the Gaussian membership function, the passenger preference weight vector P can obtain the passenger preference factor f through the fuzzy logic system, and the passenger preference weight vector with one-dimensional length is further reduced to a numerical value, and the specific process is as follows:

[0155] S2.1, define the triangular membership function:

[0156]

[0157] In the formula, a represents the left end point of the triangle in the triangular membership function, b represents the right end point of the triangle in the triangular membership function, and c represents the peak point of the triangle in the triangular membership function, x represents the input of the triangular membership function, i.e. each weight value in the passenger preference weight vector P, and μ T (x) represents the triangular membership function with input x;

[0158] S2.2, define the Gaussian membership function for defining the fuzziness:

[0159]

[0160] In the formula, μ and σ represent the mean and standard deviation of the Gaussian distribution respectively, y represents the output value of the triangular membership function, and μ G (y) represents the Gaussian membership function with input y;

[0161] S2.3, input each weight value in the passenger preference weight vector P to the triangular membership function and then pass through the Gaussian membership function, so that the final passenger preference factor f can be obtained by summation calculation; the expression is as follows:

[0162]

[0163] In the formula, the passenger preference factor f ∈ [-1, 1] directly reflects the type of the passenger, including: aggressive, ordinary and conservative, corresponding to the passenger preference factor f = -1, f = 0 and f = 1 respectively;

[0164] The above steps realize a self-supervised mapping mode of the passenger preference weight factor, overcoming the technical difficulty of requiring a large amount of calibration data in supervised learning such as deep learning;

[0165] S3, through the three-layer personalized lane changing decision model, a mapping relationship between the passenger acceptable collision time and the passenger demand factor is established, and a lane changing trajectory planning decision of the ego vehicle is generated;

[0166] The three-layer personalized lane changing decision model includes: a risk assessment layer, a global benefit assessment layer and a personalized assessment layer;

[0167] The establishment process is as follows:

[0168] S3.1, in the risk assessment layer, the artificial potential field theory is introduced, and the risk value of each lane is calculated, and the calculation formula is as follows:

[0169]

[0170] In the formula, represents the potential field size of the oth lane; r represents the number of all obstacle vehicles on the oth lane; t e represents the lane changing duration; η obs is the lane weight, and in the embodiment, η obs = 0.5; A μ is the longitudinal cross-sectional area of the μth obstacle vehicle, which is used to represent the size of the vehicle, and μ ∈ (1, +∞); is the Euclidean distance between the ego vehicle and the μth obstacle vehicle at the tth moment; Q represents the potential field influence range of the vehicle, and in the embodiment, Q = 10 m, that is, when the Euclidean distance between the ego vehicle and the obstacle vehicle exceeds Q, the potential of the vehicle on the current lane is not calculated; represents the relative speed of the ego vehicle and the μth obstacle vehicle at the tth moment, and due to the limitation of the planning frequency of 10 Hz, the interval between the tth moment and the t+1th moment is 0.1 s; It should be noted that the total number of lanes is limited to 2 in the embodiment, which can represent the basic model of lane changing, that is, o ∈ [1, 2]; The total number of perceived vehicles r is not limited, and is determined according to the specific road conditions, so r ∈ [0, +∞) and μ ∈ [0, r]; Since the potential field needs to be calculated for each trajectory point of the ego vehicle during lane changing, t ∈ [0, t e ]; In addition, o and r are integers in the interval range; The above process is true if and only if When the gap is greater than the gap threshold value ξ, the ego vehicle generates a lane change intention. In this embodiment, ξ = 0.7 is taken after calculation.

[0171] The gap threshold value is calculated as follows:

[0172] The minimum braking distance of the ego vehicle and the obstacle vehicle is obtained, and the expression is as follows:

[0173]

[0174] In the formula, a bmax The maximum braking deceleration of the ego vehicle is represented by s, and the minimum braking distance of the ego vehicle and the obstacle vehicle is represented by s, and the following can be obtained:

[0175]

[0176] That is, the potential field size of the oth lane can be regarded as a function of the minimum braking distance s, where s ∈ (0, +∞), and the function is a convex function in the domain. Considering safety, this embodiment calculates the value of the minimum braking distance s that can make the potential field of the oth lane Minimum, that is, the following equation is solved:

[0177]

[0178] In the formula, The differential of the potential field of the oth lane with respect to the minimum braking distance s of the ego vehicle and the obstacle vehicle is represented by s, and the solution of the equation is regarded as the gap threshold value ξ in this embodiment;

[0179] The potential field range is as shown in Figure 5

[0180] It should be noted that after generating the lane change intention, it is still necessary to determine whether to generate a lane change decision through subsequent calculation;

[0181] S3.2, in the global evaluation layer, a set of optimal strategies is generated between the ego vehicle and the rear vehicle of the target lane through a pure strategy game method, which can minimize the impact on the overall traffic operation. The steps of obtaining the optimal strategy are as follows:

[0182] S3.2.1, a set of strategies with the highest benefit is generated between the ego vehicle and the rear vehicle of the target lane by using a pure strategy game method, and a lane change decision is made;

[0183]

[0184] In the formula, U VC (t) and U VT ​(t) represents the benefit value of the current lane and the target lane of the ego vehicle at time t, respectively; U(t) is the total benefit of the local traffic system at time t; Γ represents the number of all obstacle vehicles in front of the current lane of the ego vehicle; p and q are the number of all obstacle vehicles in front of and behind the target lane from the perspective of the ego vehicle, respectively; ω t is the time weight value of γ at time t; V(t), and represents the vehicle, the obstacle vehicle in front of the current lane, the obstacle vehicle in front of the target lane, and the obstacle vehicle behind the target lane. The speed of the obstacle car at the t moment; where Γ∈[0,+∞), p∈[0,+∞), q∈[0,+∞), γ∈[0,+∞), ρ∈[0,+∞), The specific value depends on the perception limit of the vehicle. Therefore, the values ​​of Γ, p and q are a dynamic process.

[0185] All strategy combinations include: "the ego vehicle changes lanes and the vehicle behind it in the target lane accepts" and "the ego vehicle does not change lanes and the vehicle behind it in the target lane accepts". The global benefit value of each strategy combination is calculated based on the above global benefit function. The evaluation is completed by finding the optimal benefit combination. It is important to note that the global benefit evaluation will only continue if the ego vehicle's strategy is to change lanes. Otherwise, the decision of not changing lanes is directly output.

[0186] S3.3. If both layers have implemented lane-changing strategies, the passenger's needs are further evaluated by comparing the collision time between the ego vehicle and the vehicle behind it in the target lane to determine whether the passenger's personalized needs are met.

[0187] Specifically, in the personalized evaluation layer, a mapping relationship between the minimum acceptable collision time for passengers and the passenger demand factor f is established. The establishment process is as follows:

[0188] S3.3.1. Predicted collision duration. The expression for predicted collision duration is as follows:

[0189]

[0190] Where, T A Indicates the predicted collision time; D TF is the Euclidean distance between the ego vehicle and the vehicle behind it in the target lane; |vv TF | represents the absolute difference in speed between the ego vehicle and the vehicle behind it in the target lane; v represents the speed of the ego vehicle; v TF is the speed of the vehicle behind in the target lane;

[0191] S3.3.2. Establish a mapping relationship between the passenger's acceptable collision time and the passenger demand factor f, as shown in the following expression:

[0192] Where, T P Indicates the acceptable collision time for passengers; T S is a time constant, which is 3 seconds in this embodiment; T Indicates the scaling weight, which is set to 0.3 based on experience;

[0193] S3.3.3. Obtain the final output by predicting the collision duration and mapping the relationship;

[0194] Specifically, based on the above analysis, when T A >T P When , it indicates that the predicted collision time with the vehicle behind in the target lane meets the passenger's needs, that is, the final output is a decision to allow the lane change; otherwise, it indicates that the passenger rejects the lane change decision, and the lane change decision is not executed;

[0195] S4. By establishing a trajectory generation model based on a multidimensional coordinate space, the lane-changing trajectory planning decision problem of the ego vehicle is transformed into a system of multivariate linear equations for solving the lane-changing trajectory; the expression is as follows:

[0196]

[0197] in, as well as They are respectively represented as the initial time t of the lane change of the vehicle s The projected distance on the centerline of the target lane, the initial speed, the initial acceleration, the initial plane horizontal coordinate, the initial plane vertical coordinate, and the slope of the path point at the initial moment. It should be noted that in this embodiment, t s =0; as well as They represent the projected distance of the vehicle on the centerline of the target lane at the end of the lane change, the speed at the end, the acceleration at the end, the horizontal coordinate at the end, the vertical coordinate at the end, and the end time t e The slope of the path point; for the convenience of calculation, the above state quantities are combined into the initial state vector and the final state vector When the two state vectors are determined, the trajectory polynomial coefficient vectors α and β can also be determined by solving a multivariate linear equation system. Since all state quantities in the initial state vector can be obtained by the vehicle sensors, it is necessary to further calculate the state quantities in the final state vector. If you notice that the target lane centerline is actually a sequence of discrete points, you only need to search for the point with the same value through the Euclidean distance. The nearest point, and through linear interpolation between the previous point and the next point, we can solve all the points except the end time t e , speed at the end and acceleration all state quantities of the ego vehicle;

[0198] S5, solving the unknowns of the lane-changing trajectory multivariate linear equation set by optimization;

[0199] That is, by setting a cost function and constraint condition considering passenger preferences, the unknown quantity solving problem of the lane-changing trajectory multivariate linear equation set is converted into a nonlinear optimization problem; the specific steps are as follows:

[0200] S5.1, establishing a multi-objective cost function considering safety, comfort and efficiency according to passenger preferences, including the following steps:

[0201] S5.1.1, introducing a lane-level potential field as a safety constraint, which comprehensively considers relative speed, acceleration and longitudinal cross-sectional area of the vehicle along the y direction;

[0202] The lane-level potential field calculation expression is as follows:

[0203]

[0204] In the formula, and respectively represent the corresponding distance values of the projection of the ego vehicle, the obstacle vehicle behind the ego vehicle in the target lane, the obstacle vehicle in front of the ego vehicle in the target lane and the obstacle vehicle in front of the ego vehicle in the current lane on the lane-changing target lane at the t-th moment; and respectively represent the speeds of the ego vehicle and the obstacle vehicle behind the ego vehicle in the target lane at the t-th moment; and are the maximum braking decelerations of the ego vehicle and the obstacle vehicle behind the ego vehicle in the target lane, and in the embodiment, g is taken τ is the planning time step of the ego vehicle, and in the embodiment, τ = 0.1 s is taken; Q is the potential energy influence range of each vehicle, and in the embodiment, Q = 10 m is taken;

[0205] S5.1.2, the safety cost function is satisfied by performing second-order Taylor expansion at each trajectory point to meet the quadratic form; the expression is as follows:

[0206]

[0207]

[0208] In the formula, Js represents the safety cost function; X represents a high-dimensional unknown form; X0 is the input of each trajectory point; is the gradient calculation; H represents the Hessian matrix; represents the partial differential calculation;

[0209] S5.1.3, introduce the acceleration and jerk of each trajectory point of the ego vehicle on the comfort and high efficiency cost function to characterize; the expression is as follows:

[0210]

[0211] In the formula, J c and J e respectively represent the comfort cost function and the high efficiency cost function; j(t) represents the jerk of the ego vehicle at t time; t e represents the total length of the lane change;

[0212] S5.1.4, obtain a multi-objective cost function combining safety, comfort and high efficiency; the expression is as follows:

[0213] J=w s ·J s +w c ·J c +w e ·J e

[0214] In the formula, w s , w c and w e are three index weight values in the passenger preference weight vector P;

[0215] S5.2, optimize the hyperparameters of the long short-term memory neural network through the whale optimization algorithm to obtain the upper and lower limits of the safety constraint interval;

[0216] Specifically, the optimized neural network is trained through an open source data set to output the lane change duration prediction value of the ego vehicle at the lane change starting time, and the lane change duration constraint is obtained in combination with the passenger riding preference factor f; the safety constraint is linearized to improve the solving speed of the model;

[0217] The specific steps are as follows:

[0218] S5.2.1, establish a long short-term memory neural network based on whale optimization algorithm optimization to obtain the estimated driving distance; the expression is as follows:

[0219]

[0220] In the formula, is the speed of the ego vehicle at the initial lane change time, is the estimated driving distance, and respectively represent the speed of the ego vehicle and the target lane in front of the obstacle vehicle;

[0221] S5.2.2, 400 groups of data are collected from the NGSIM open source dataset, 70% of which are randomly selected as the training set, and the remaining 30% are used as the test set; and the whale optimization algorithm is used to optimize the hyperparameters of the long short-term memory neural network, including: learning rate and maximum iteration number;

[0222] The loss value is calculated by the loss function, and the whale optimization algorithm needs to search for the value of the hyperparameter when the loss value is minimized to update the model weight in reverse:

[0223]

[0224] In the formula, Λ represents the total number of training samples, and here Λ=350; T φ and respectively represent the real lane changing time and the model predicted lane changing time; F represents the prediction loss value of each iteration, that is, the fitness function value of the whale optimization algorithm;

[0225] The final training result is shown in Figure 6 It can be seen that the lane changing time prediction result of the long short-term memory neural network optimized by the whale optimization algorithm is closer to the real value than that of the ordinary long short-term memory neural network; the fitness function value of the whale optimization algorithm reaches the minimum and converges when it is iterated to the 10th generation.

[0226] S5.2.3, the passenger riding preference factor f is introduced, and the following lane changing time constraint form is designed:

[0227] t e ∈[T min ,T max ]

[0228]

[0229] In the formula, T min represents the lower bound of the lane changing time, T max represents the upper bound of the lane changing time, represents the predicted lane changing time output by the model;

[0230] S5.2.4, collision constraints are introduced during vehicle driving to ensure that the vehicle does not collide;

[0231] Specifically, during vehicle driving, it is necessary to ensure that the vehicle does not collide, so it is necessary to introduce collision constraints to the optimization problem; in order to improve the calculation efficiency of the model, the collision constraint of the embodiment can be expressed in a linear form;

[0232] Specifically, as shown in Figure 7 , the distance of the midpoint of the rear axle of the ego vehicle from the initial lane changing time to the projection on the target lane is Then, the three vertices of the rear bumper left point, the front bumper left point and the front bumper right point of the ego vehicle can be calculated by geometric calculation, as follows:

[0233]

[0234] where w ego is the vehicle width of the ego vehicle, and w ego = 2m in the embodiment; θ t is the heading angle of the ego vehicle at time t, and θ t ∈(0, 2π); lr ego represents the length from the midpoint of the rear axle of the ego vehicle to the rear bumper of the ego vehicle, and lr ego = 1m in the embodiment; and respectively represent the projection distance values of the rear bumper left point, the front bumper left point and the front bumper right point of the ego vehicle on the target lane center line at time t, and the value ranges of and are both [0, +∞); and and are combined into a vector S Ω , and it is obvious that S Ω is the projection distance value of the ego vehicle on the target lane center line which is a linear function of χ∈{A, B, C}; at this time, the ego vehicle safety linearization constraint can be expressed as:

[0235]

[0236] where f colli and r colli respectively represent the safety distance threshold in front of the ego vehicle and the safety distance threshold behind the ego vehicle, and f colli = r colli = 2m in the embodiment; s 1,t , s 2,t and s 3,t respectively represent the projection distance value of the target lane rear obstacle vehicle on the target lane, the projection distance value of the target lane front obstacle vehicle on the target lane and the projection distance value of the target lane front obstacle vehicle on the target lane of the ego vehicle currently located in the lane;

[0237] S5.2.5. The planned trajectory needs to apply a motion state constraint at the end of the lane changing;

[0238] Specifically, in order to ensure that the ego vehicle can smoothly follow the traffic flow of the target lane at the end of the lane changing, the motion state constraint needs to be applied to the end point of the lane changing trajectory planned by the ego vehicle;

[0239] The embodiment introduces a vehicle following model to calculate the motion state of the lane-changing endpoint, and derives the acceleration constraint of the lane-changing endpoint time as:

[0240]

[0241] In the formula, represents the vector composed of the accelerations of the obstacle vehicles located behind and in front of the ego vehicle on the target lane; is the speed of the obstacle vehicle in front of the current lane of the ego vehicle at the lane-changing endpoint; Δs * is a distance constant, and the value range is Δs * ∈ [1, 5], and the embodiment takes Δs * = 2; ω1 and ω2 are the distance weight value and the relative speed weight value respectively, the distance weight value value range is ω1 ∈ [0, 1], and the embodiment takes the distance weight value as ω1 = 1; the relative speed weight value value range is ω2 ∈ [0, 1], and the embodiment takes the relative speed weight value as ω2 = 1; l OV-3 and l ego respectively represent the length of the obstacle vehicle in front of the current lane of the ego vehicle and the length of the ego vehicle; therefore, the motion state constraint can be summarized as:

[0242]

[0243] In the formula, represents the projection distance of the ego vehicle on the center line of the target lane at the t-th time; represents the speed of the ego vehicle at the t-th time;

[0244] S5.3, the trajectory optimization problem described in the above steps S5.1 and S5.2 is arranged into a nonlinear programming problem, and the complete expression is as follows:

[0245]

[0246] The expression of the above nonlinear programming problem is brought into the IPOPT sequence quadratic programming solver to solve, and the personalized planning trajectory is obtained;

[0247] S6, in order to verify the effect of the trajectory planning model, the embodiment designs three passengers with different preference types, which are conservative passenger 1, ordinary passenger 2 and aggressive passenger 3; the embodiment carries out simulation experiments on the experimental parameter configuration as shown in Table 1, and the simulation road is a set of randomly generated smooth discrete points:

[0248] Table 1 Simulation experiment specific parameter table

[0249]

[0250] The lane-changing trajectory planning results of different test scenarios are as followsFigure 8 as shown in FIG. 8(a);

[0251] The lane-changing trajectory planning method designed in this embodiment plans motion curves under three different passengers. In this scenario, all three passengers tend to perform lane-changing for the ego vehicle, as shown in FIG. 8(a). The aggressive passenger plans a shorter lane-changing path and adopts a higher speed and acceleration to achieve the minimum lane-changing duration, as shown in FIG. 8(b). The conservative passenger plans a longer lane-changing path and adopts a lower speed and acceleration to achieve the maximum lane-changing duration, as shown in FIG. 8(c). The normal passenger plans a medium lane-changing path and adopts a medium speed and acceleration to achieve a medium lane-changing duration, as shown in FIG. 8(d). Thus, it can be seen that the lane-changing trajectory planning method proposed in this embodiment realizes personalized lane-changing trajectory planning by emphasizing differentiated lane-changing decision sensitivity, different passenger experiences, and the balance between driving efficiency and comfort. Figure 8 Figure 8

[0252] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.​​

Claims

1. A lane-changing trajectory planning method for an intelligent connected vehicle that considers the personalized riding needs of passengers, characterized by: The following steps are involved: S1. Convert the passenger's riding preference into a judgment matrix, and map the judgment matrix into a passenger preference weight vector; Passengers' riding preferences include: riding comfort preference, safety preference and efficiency preference; The preference degree of riding comfort, safety and efficiency is formed into a judgment matrix A in the form of a 3×3 matrix; wherein the element A in the i-th (i=1, 2, 3) row and the j-th (j=1, 2, 3) column is ij It represents the interaction between passengers’ perception of ride comfort, safety and efficiency; S2. Establish a secondary mapping method to obtain the passenger preference factor through the passenger preference weight vector; The quadratic mapping method includes: triangular membership function and Gaussian membership function; The steps of establishing the secondary mapping method and obtaining the passenger preference factor through the passenger preference weight vector are as follows: S2.

1. Define the triangular membership function: Where a represents the left endpoint of the triangle in the triangular membership function, b represents the right endpoint of the triangle in the triangular membership function, and c represents the peak point of the triangle in the triangular membership function. x represents the input of the triangular membership function, that is, the weight values ​​in the passenger preference weight vector P. μ T (x) represents the triangular membership function with input x; S2.

2. Define the Gaussian membership function to define fuzziness: Where μ and σ represent the mean and standard deviation of Gaussian distribution respectively, y represents the output value of triangular membership function, μ G (y) represents the Gaussian membership function with input y; S2.

3. Input each weight value in the passenger preference weight vector P into the triangular membership function and then pass it through the Gaussian membership function. The final passenger preference factor f can be obtained by summing the values. The expression is as follows: Wherein, the passenger preference factor f∈[-1,1] directly reflects the type of passenger, including: radical, ordinary, and conservative, corresponding to passenger preference factors f=-1, f=0, and f=1, respectively; S3. Using a three-layer personalized lane-changing decision model, a mapping relationship is established between the passenger's acceptable collision time and the passenger demand factor, generating a lane-changing trajectory planning decision for the ego vehicle. The three-layer personalized lane-changing decision model includes: a risk assessment layer, a global benefit assessment layer, and a personalized assessment layer; S4. By establishing a trajectory generation model based on a multidimensional coordinate space, the lane-changing trajectory planning decision problem of the ego vehicle is transformed into solving a multivariate linear equation system for the lane-changing trajectory; S5. Optimize and solve the unknown variables of the multivariate linear equations of the lane-changing trajectory to complete the lane-changing of the vehicle.

2. The method for planning lane-changing trajectories for an intelligent connected vehicle that considers individual passenger needs according to claim 1, characterized in that: The method for solving the passenger preference weight vector is the hierarchical analysis method, and the specific steps include: S1.

1. Calculate the sum of each column of the judgment matrix; the expression is as follows: Where S j represents the sum of the elements in the jth column of the judgment matrix A; A ij Represents the element in row i and column j of the judgment matrix A, Indicates the sum of the elements in the jth column; S1.

2. Standardize each element in the judgment matrix A; the expression is as follows: Where A' ij Represents the standardized judgment matrix elements; S1.

3. Calculate each weight value in the passenger preference weight vector; the expression is as follows: Where P represents the passenger preference weight vector; P k represents the kth element in the passenger preference weight vector; S1.

4. Ensure the rationality of the judgment matrix by performing a consistency check on the customer preference weight vector. The specific steps are as follows: S1.4.

1. Calculate the maximum eigenvalue; the expression is as follows: Where λ max is the maximum eigenvalue of the judgment matrix A, (A·P) k The kth element of the product of the judgment matrix A and the passenger preference weight vector P represents the comprehensive score of each factor; S1.4.

2. Calculate the consistency index; the expression is as follows: Where CI represents the consistency index value, and n represents the matrix dimension.

3. The method for planning lane-changing trajectories for an intelligent connected vehicle that considers individual passenger needs according to claim 1, characterized in that: The steps of establishing a mapping relationship between the passenger's acceptable collision time and the passenger demand factor through the three-layer personalized lane change decision model and generating the lane change trajectory planning decision of the ego vehicle are as follows: S3.

1. In the risk assessment layer, artificial potential field theory is introduced to calculate the risk value of each lane. The calculation formula is as follows: Where, represents the potential energy field size of lane o; r represents the number of all obstacle vehicles on lane o; t e Indicates the duration of lane change; η obs is the lane weight; A μ is the longitudinal cross-sectional area of ​​the μth obstacle vehicle, which is used to characterize the size of the vehicle, μ∈(1,+∞); is the Euclidean distance between the ego vehicle and the μth obstacle vehicle at time t; Q represents the influence range of the vehicle's potential energy field, that is, when the Euclidean distance between an obstacle vehicle and the ego vehicle exceeds Q, the potential energy of the vehicle in the current lane is not calculated; represents the relative speed between the ego vehicle and the μth obstacle vehicle at time t. Due to the limitation of 10 Hz planning frequency, the interval between time t and time t+1 is 0.1s. In addition, o and r are integers within the interval range. The above process is valid only if When the gap is greater than the threshold ξ, the ego vehicle has the intention to change lanes; The calculation process of the gap threshold ξ is as follows: Get the minimum braking distance between the vehicle and the obstacle vehicle without collision; the expression is as follows: Where, a bmax represents the maximum braking deceleration of the ego vehicle, and s represents the minimum braking distance between the ego vehicle and the obstacle vehicle without collision. At this time, we can obtain: That is, the potential energy field of lane o can be regarded as a function of the minimum braking distance s, where s∈(0,+∞). The function is a convex function in the domain of definition. When the minimum braking distance s takes any value, the potential energy field of lane o can be Minimize, that is, solve the following equation: Where, represents the differential of the potential energy field of lane o with respect to the minimum braking distance s for the ego vehicle to avoid collision with the obstacle vehicle. The solution of the equation is the gap threshold ξ; S3.

2. In the global evaluation layer, a pure strategy game method is used to generate an optimal strategy between the ego vehicle and the vehicle behind it in the target lane, which can minimize the impact on the overall traffic situation. The steps to obtain the optimal strategy are as follows: S3.2.

1. Generate a set of strategies with the highest efficiency between the ego vehicle and the vehicle behind it in the target lane using a pure strategy game method, and make a lane change decision. The global benefit function expression is: Where U VC (t) and U VT (t) represents the benefit value of the current lane and the target lane of the ego vehicle at time t, respectively; U(t) is the total benefit of the local traffic system at time t; Γ represents the number of all obstacle vehicles in front of the current lane of the ego vehicle; p and q are the number of all obstacle vehicles in front of and behind the target lane from the perspective of the ego vehicle, respectively; ω t is the time weight value at time t; V(t), and represents the vehicle, the γth obstacle vehicle in front of the current lane, the ρth obstacle vehicle in front of the target lane, and the ρth obstacle vehicle behind the target lane. The speed of the obstacle car at the t moment; where Γ∈[0,+∞), p∈[0,+∞), q∈[0,+∞), γ∈[0,+∞), ρ∈[0,+∞), The specific value depends on the perception limit of the vehicle. Therefore, the values ​​of Γ, p and q are a dynamic process. All strategy combinations include: "the ego vehicle changes lanes and the vehicle behind it in the target lane accepts" and "the ego vehicle does not change lanes and the vehicle behind it in the target lane accepts". The global benefit value of each strategy combination is calculated based on the above global benefit function, and the evaluation is completed by finding the optimal benefit combination. S3.

3. If both layers have implemented lane-changing strategies, the passenger's needs are further evaluated by comparing the collision time between the ego vehicle and the vehicle behind it in the target lane to determine whether the passenger's personalized needs are met. The steps are as follows: S3.3.

1. Predicted collision duration. The expression for predicted collision duration is as follows: Where, T A Indicates the predicted collision time; D TF is the Euclidean distance between the ego vehicle and the vehicle behind it in the target lane; |vv TF | represents the absolute difference in speed between the ego vehicle and the vehicle behind it in the target lane; v represents the speed of the ego vehicle; v TF is the speed of the vehicle behind in the target lane; S3.3.

2. Establish a mapping relationship between the passenger's acceptable collision time and the passenger demand factor f, as shown in the following expression: T P =T S -η T ·f Where, T P Indicates the acceptable collision time for passengers; T S is a time constant, take 3 seconds; η T Indicates the scaling weight, with a value of 0.3; S3.3.

3. Obtain the final output by predicting the collision duration and mapping the relationship; The final output is when T A >T P When the predicted collision time with the vehicle behind in the target lane meets the passenger's needs, the final output is a decision to allow the lane change; otherwise, it indicates that the passenger rejects the lane change decision, and the lane change decision is not executed.

4. The method for planning lane-changing trajectories for an intelligent connected vehicle that considers individual passenger needs according to claim 1, characterized in that: The expression of the lane-changing trajectory multivariate linear equation group is as follows: in, as well as They are respectively represented as the initial time t of the lane change of the vehicle s The projected distance on the centerline of the target lane, the initial speed, the initial acceleration, the initial plane horizontal coordinate, the initial plane vertical coordinate, and the slope of the path point at the initial time; as well as They represent the projected distance of the vehicle on the centerline of the target lane at the end of the lane change, the speed at the end, the acceleration at the end, the horizontal coordinate at the end, the vertical coordinate at the end, and the end time t e The slope of the path point; for the convenience of calculation, the above state quantities are combined into the initial state vector and the final state vector When the two state vectors are determined, the trajectory polynomial coefficient vectors α and β can also be determined by solving a multivariate linear equation system. Since all state quantities in the initial state vector can be obtained through the vehicle-mounted sensors, it is necessary to further calculate the state quantities in the final state vector; When the target lane centerline is a discrete point sequence, the Euclidean distance is used to search for the point that matches the target lane centerline. The nearest point, and through linear interpolation between the previous point and the next point, we can solve all the points except the end time t e , speed at the end and acceleration All state quantities.

5. The method for planning lane-changing trajectories for an intelligent connected vehicle that considers individual passenger needs according to claim 1, characterized in that: The steps of the optimization solution are as follows: S5.

1. Establish a multi-objective cost function that considers safety, comfort, and efficiency based on passenger preferences, including the following steps: S5.1.

1. A lane-level potential energy field is introduced as a safety constraint, taking into account relative velocity, acceleration, and the longitudinal cross-sectional area of ​​the vehicle in the y direction. The lane-level potential energy field calculation expression is as follows: Where, as well as They represent the projection distances of the ego vehicle, the obstacle vehicle in the target lane behind the ego vehicle, the obstacle vehicle in the target lane in front of the ego vehicle, and the obstacle vehicle in the current lane in front of the ego vehicle on the target lane at time t. as well as They represent the speed of the ego vehicle and the obstacle vehicle in the target lane behind the ego vehicle at time t respectively; and is the maximum braking deceleration of the ego vehicle and the obstacle vehicle in the target lane behind the ego vehicle; τ is the planning time step of the ego vehicle; Q is the potential energy influence range of each vehicle; S5.1.

2. The safety cost function is expanded to satisfy the quadratic form by performing a second-order Taylor expansion at each trajectory point; the expression is as follows: Where, J s represents the security cost function; X represents the form of high-dimensional unknowns; X0 is the input of each trajectory point; ▽ is the gradient calculation; H represents the Hessian matrix; represents partial differential calculation; S5.1.

3. In the comfort and efficiency cost functions, the acceleration and jerk of each trajectory point are introduced to represent the ego vehicle. The expressions are as follows: Where, J c and J e They represent the comfort cost function and the efficiency cost function respectively; j(t) represents the jerk of the vehicle at time t; t e Indicates the total lane change time; S5.1.

4. Combine safety, comfort, and efficiency to derive a multi-objective cost function; the expression is as follows: J=w s ·J s +w c ·J c +w e ·J e Where w s 、w c and w e are the weight values ​​of the three indicators in the passenger preference weight vector P; S5.

2. Optimize the hyperparameters of the long short-term memory neural network using the whale optimization algorithm to obtain the upper and lower limits of the safety constraint. The specific steps are as follows: S5.2.

1. Build a long short-term memory neural network optimized by the whale optimization algorithm to obtain the estimated driving distance; the expression is as follows: Where, is the speed of the vehicle at the initial moment of lane change, is the estimated driving distance, Indicates the speed of the obstacle vehicle in front of the ego vehicle in the target lane; S5.2.

2. Collect a dataset from an open source dataset, randomly select 70% of it as a training set and the remaining 30% as a test set; and use the whale optimization algorithm to optimize the hyperparameters of the long short-term memory neural network, including the learning rate and the maximum number of iterations; By calculating the loss value through the loss function, we can determine the value of the hyperparameters that the whale optimization algorithm needs to search for to minimize the loss value, and then update the model weights in reverse: Where Λ represents the total number of training samples; T φ and They represent the actual lane change time and the lane change time predicted by the model respectively; F represents the prediction loss value of each iteration, that is, the fitness function value of the whale optimization algorithm; S5.2.

3. Introduce the passenger preference factor f and design the following lane change time constraint: Where, T min represents the lower bound of the lane change time, T max represents the upper bound of the lane change duration, Indicates the predicted lane change duration output by the model; S5.2.

4. Introduce collision constraints during vehicle movement to ensure that the vehicle does not collide; Collision constraints are expressed in linearized form; The distance projected on the target lane by the midpoint of the rear axle of the ego vehicle at the initial moment of lane change is The three vertices, namely the left side of the rear bumper, the left side of the front bumper, and the right side of the front bumper, can be calculated through geometric calculation as follows: Where w ego is the width of the vehicle; θ t is the heading angle of the vehicle at time t, θ t ∈(0,2π);lr ego Indicates the length from the midpoint of the vehicle's rear axle to the vehicle's rear bumper; and They represent the projection distance values ​​of the left point of the rear bumper, the left point of the front bumper, and the right point of the front bumper on the center line of the target lane at time t, respectively. and The value range of is [0,+∞); and Combined into a vector S Ω , then obviously S Ω is the projection value of the vehicle's distance on the center line of the target lane A linear function of χ∈{A,B,C}; in this case, the linearization constraint of the ego-vehicle safety can be expressed as: Where, f colli and r colli They represent the safety distance threshold in front of the vehicle and the safety distance threshold behind the vehicle respectively; s 1,t 、s 2,t and s 3,t They represent the projection value of the obstacle vehicle behind the target lane on the target lane, the projection value of the obstacle vehicle in front of the target lane on the target lane, and the projection value of the obstacle vehicle in front of the lane where the vehicle is currently located on the target lane; S5.2.5, The planned trajectory needs to impose motion state constraints at the end of the lane change; The way to impose motion state constraints is to introduce a vehicle following model to calculate the motion state at the lane change endpoint, and derive the acceleration constraint at the lane change endpoint as: Where, The vector representing the acceleration of the obstacle vehicles behind and in front of the vehicle in the target lane; Δs is the speed of the obstacle vehicle in front of the vehicle in the current lane at the end of the lane change; * Is a distance constant with a value range of Δs * ∈[1,5]; ω1 and ω2 are the vehicle distance weight value and relative speed weight value respectively. The vehicle distance weight value range is ω1∈[0,1]; the relative speed weight value range is ω2∈[0,1]; l OV-3 and l ego They represent the length of the obstacle vehicle in front of the current lane and the length of the vehicle respectively; therefore, the motion state constraints can be summarized as: Where, represents the projection distance of the vehicle on the center line of the target lane at time t; represents the speed of the vehicle at time t; S5.

3. The trajectory optimization problem described in steps S5.1 and S5.2 is transformed into a nonlinear programming problem. The complete expression is as follows: Substituting the expression of the above nonlinear programming problem into the IPOPT sequential quadratic programming solver, the personalized planning trajectory can be obtained.

Images