An automatic driving variable neighborhood search and trajectory planning method for urban scenes

Abstract

The application discloses a kind of automatic driving variable neighborhood search and trajectory planning method for urban scene, involve automatic driving control technical field, comprising the following steps: according to surrounding environment perception and high-precision positioning device, vehicle is positioned;Mathematical description is carried out to rectangle and trapezoidal neighborhood;Trajectory planning and optimization are carried out to the search of independent interval neighborhood;The continuity and stability of trajectory under variable neighborhood search are guaranteed using average residence time method;The application is based on variable neighborhood search under the premise, according to traffic sign under the complex road scene of city, the trajectory planning of vehicle driving is carried out in combination with the dynamics constraint of vehicle, and trajectory planning is divided into single neighborhood planning and complete trajectory planning under the complex road scene of city, the stability of different trajectory switching is guaranteed in complete trajectory planning, so as to improve the differential information capture ability of intelligent car facing the complex scene of city intersection and the planning and control ability of path, improve the driving safety of intelligent car.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving control technology, specifically to an autonomous driving variable neighborhood search and trajectory planning method for urban scenarios. Background Technology

[0002] With the acceleration of urbanization, urban traffic management faces unprecedented challenges. The increasing complexity of urban transportation networks, coupled with massive and varied traffic flows and the interplay of pedestrians, vehicles, and non-motorized vehicles, makes traffic management exceptionally complex. This complexity not only affects travel efficiency but also increases the risk of traffic accidents, threatening the property and lives of urban residents. The Ministry of Transport's implementation of the "Urban Public Transportation Regulations" aims to improve urban traffic management through intelligent means. This invention, specifically addressing the unique layout and traffic flow of urban scenarios, proposes an autonomous driving variable neighborhood search and trajectory planning method for urban environments. This method aims to solve the problems of insufficient differential information capture, poor path planning and control, and high accident risk in current intelligent vehicles facing complex scenarios such as urban intersections. Summary of the Invention

[0003] The purpose of this invention is to provide a method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios, so as to solve the problems existing in the prior art mentioned in the background.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios includes the following steps:

[0006] S1: The vehicle is located based on the surrounding environment perception and high-precision positioning device;

[0007] S2: Provide a mathematical description of the rectangular and trapezoidal neighborhoods;

[0008] S3: Perform trajectory planning and optimization for the search of independent interval neighborhoods;

[0009] S4: The average dwell time method is used to ensure the continuity and stability of the trajectory under the variable neighborhood search.

[0010] Preferably, in step S1, locating the vehicle includes dividing the vehicle into various positions during its journey through complex urban environments, and identifying the trigger point for the vehicle's neighborhood search transition.

[0011] Preferably, in S2, the mathematical description of the neighborhood facilitates the division of the vehicle's perceived environment, and the complete trajectory is planned by superimposing the divisions of different neighborhoods.

[0012] Preferably, in S3, trajectory planning and optimization for searching independent interval neighborhoods includes planning within rectangular search neighborhoods and planning within trapezoidal search neighborhoods. In each interval of neighborhood switching search, the surrounding environment is segmented and superimposed to obtain regular shapes such as rectangles and trapezoids, thereby planning a complete trajectory.

[0013] Preferably, in step S4, ensuring the stability and continuity of the complete trajectory involves the following steps:

[0014] S41: Use the Lyapunov function to determine the stability of the trajectory within a single neighborhood and its convergence time;

[0015] S42: Use the average dwell time method to ensure the stability of neighborhood search switching.

[0016] Preferably, the mathematical description of the trapezoidal search neighborhood is as follows:

[0017] The mathematical expression for the trapezoidal search neighborhood is given by O(x0,y0), which represents the initial position of the intelligent vehicle. The side edges l1 and l2 of the neighborhood are both of length r, the top edge is l3, the height is s, and the bottom edge is d, as shown in the following formula:

[0018]

[0019] Trajectory planning within a trapezoidal neighborhood: Determine the starting and ending points within the trapezoidal search neighborhood, use quintic interpolation polynomial neighborhood planning to obtain a candidate trajectory set, and then obtain the optimal trajectory curve.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] This invention, based on variable neighborhood search, combines traffic signs in complex urban road scenarios with vehicle dynamic constraints to plan vehicle trajectory. The trajectory planning is divided into planning within a single neighborhood and complete trajectory planning in complex urban road scenarios. The stability of switching between different trajectories is ensured in the complete trajectory planning, thereby improving the intelligent vehicle's ability to capture differential information and plan and control paths in complex urban intersection scenarios, and improving the driving safety of the intelligent vehicle. Attached Figure Description

[0022] Figure 1 This is a schematic diagram illustrating the urban intersection environment positioning and division according to the present invention.

[0023] Figure 2 The vehicle coordinate system diagram established for this invention. Detailed Implementation

[0024] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0025] Please see Figure 1-2 The present invention provides the following technical solutions:

[0026] A method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios includes the following steps:

[0027] S1: The vehicle is located based on the surrounding environment perception and high-precision positioning device; the vehicle location includes dividing the vehicle into various positions during the driving process in complex urban scenarios, and identifying the trigger point for the vehicle's neighborhood search and conversion.

[0028] The location divisions include: the straight section before entering the urban intersection, the waiting area at the urban intersection, the driving area in the center of the urban intersection, and the exit area from the urban intersection.

[0029] Urban intersection waiting area: When a vehicle is about to enter an urban intersection, the area where the vehicle can travel in a straight line from the stop line of the zebra crossing at the urban intersection is called the urban intersection waiting area.

[0030] Central driving area of ​​urban intersections: The driving area from when a vehicle leaves the stop line and enters the core area of ​​an urban intersection according to traffic signals, until it leaves the stop line of the urban intersection, is called the central driving area of ​​urban intersections.

[0031] Urban intersection departure zone: The driving range of a vehicle from the stop line of an urban intersection to the point of leaving the designated urban intersection scenario is called the urban intersection departure zone.

[0032] After defining the urban intersection environment, the parking line leading to the intersection is used as the switching point for trapezoidal neighborhood search; similarly, the parking line is used as the switching point for rectangular neighborhood search when a vehicle leaves the intersection. Figure 1 As shown, the black dashed lines in the intersection are the stop lines for vehicles, corresponding to the zebra crossings at the city intersection. This means that when a vehicle passes through a city intersection, it needs to switch between two neighborhood searches: first, from rectangular neighborhood search to trapezoidal neighborhood search, and second, from trapezoidal neighborhood search to rectangular neighborhood search.

[0033] S2: Mathematically describe rectangular and trapezoidal neighborhoods; the mathematical description of the neighborhood facilitates the division of the vehicle's perceived environment. The complete trajectory is planned by superimposing the divisions of different neighborhoods. The mathematical description of the neighborhood facilitates the division of the vehicle's surrounding environment and the superposition of neighborhoods to form a single neighborhood, realizing the planning of the trajectory within an independent neighborhood.

[0034] As attached Figure 2 As shown, to combine vehicle movement with the description of rectangular and trapezoidal neighborhood searches, a vehicle coordinate system needs to be established, with the front axle of the vehicle as the x-axis and the vehicle's forward direction as the y-axis. The mathematical description of the neighborhood is closely related to the superposition of trajectory searches within the neighborhood. The specific mathematical description is as follows:

[0035] Mathematical description using trapezoidal search neighborhood as an example:

[0036] The mathematical expression for the trapezoidal search neighborhood is given by O(x0,y0), which represents the initial position of the intelligent vehicle. The side edges l1 and l2 of the neighborhood are both of length r, the top edge is l3, the height is s, and the bottom edge is d, as shown in the following formula:

[0037]

[0038] Trajectory planning within a trapezoidal neighborhood: Determine the starting and ending points within the trapezoidal search neighborhood, use quintic interpolation polynomial neighborhood planning to obtain a candidate trajectory set, and then obtain the optimal trajectory curve.

[0039] S3: Trajectory planning and optimization for the search of independent interval neighborhoods; Trajectory planning and optimization for the search of independent interval neighborhoods includes planning within rectangular search neighborhoods and planning within trapezoidal search neighborhoods. In each interval of the neighborhood switching search, the surrounding environment is segmented and superimposed to obtain regular shapes such as rectangles and trapezoids, thereby planning a complete trajectory.

[0040] Trajectory planning for neighborhood search within an independent interval involves first determining the size of the neighborhood of a single interval, and then determining the start and end points of the independent interval based on the vehicle size. Taking the center of the front axle of a conventional vehicle as the start point and the start point of the next interval as the end point, the driving route within the neighborhood of a single interval is planned based on a fifth-order interpolation polynomial.

[0041] First, a sliding surface function S is designed to track and control the neighborhood search trajectory. Then, a Lyapunov function is designed to determine the stability of the neighborhood search system and the convergence time t of the neighborhood. s .

[0042] Based on the defined yaw angle tracking error and its rate of change, design the sliding surface function S.

[0043]

[0044] In Equation 1, ε i For yaw rate tracking error, For the horizontal swing angle, To anticipate the lateral angle;

[0045]

[0046] In Equation 3, β > 0, q1 and q2 are positive odd numbers, and 1 < q1 / q2 < 2; the control law U can be obtained by differentiating the synovial surface function.

[0047] We need to define the Lyapunov function V1:

[0048]

[0049] The convergence of the control system and the convergence time of the system on the sliding surface can be proven.

[0050] Redefine the Lyapunov function V2;

[0051]

[0052] In Equation 5, e i To summarize the overall error, we can obtain the convergence time t by differentiating, simplifying, and integrating it. s have to;

[0053]

[0054] In Equation 6 α > 0, v1(0) is the initial energy of the system, and p1 and p2 are positive odd numbers.

[0055] S4: The average dwell time method is used to ensure the continuity and stability of the trajectory under the variable neighborhood search; ensuring the stability and continuity of the complete trajectory is divided into the following steps:

[0056] S41: Use the Lyapunov function to determine the stability of the trajectory within a single neighborhood and its convergence time;

[0057] S42: Use the average dwell time method to ensure the stability of neighborhood search switching.

[0058] The search of a single neighborhood will not get stuck in a local optimum, ensuring the non-convexity of the trajectory; the time interval between two neighborhood searches during vehicle travel is greater than the average dwell time, ensuring the stability of the entire system when the vehicle is traveling at an urban intersection.

[0059] The core technology for solving the variable neighborhood search trajectory lies in how to address the discontinuity during the switching between different neighborhoods and how the search for the next neighborhood can get trapped in a local optimum during the switching process. To address the discontinuity during the switching between different neighborhoods, the average dwell time method in the switching system is introduced. This ensures that the interval between any two consecutive switching events is greater than the average dwell time, thus guaranteeing the continuity and exponential stability of the switching system. Define N... σ (t, τ) represents the number of handovers within the time period 0 ≤ τ ≤ t, and Save(τ) D N(0) are all switches that satisfy the following inequality:

[0060]

[0061] Where τ D It is the average dwell time, which refers to the average dwell time of the system under a neighborhood search, and N0 is the amplitude of the switching signal; as long as τ D The set is large enough that any number of transitions within it can guarantee the exponential stability of the switching system. The number of transitions at city intersections is generally only two; in special cases, meeting the above conditions is sufficient to guarantee the stability of the switching trajectory.

[0062] When switching neighborhoods, if the search for the next neighborhood gets stuck in a local optimum, the convergence time t of the neighborhood search within a single interval is calculated according to step S3 above. s When the vehicle gives a switching signal, ensure the dwell time τ. D Greater than the convergence time, i.e., τ D ≥t s This ensures that the next neighborhood search interval will not get stuck in a local optimum when switching neighborhood search intervals.

[0063] The time interval between two consecutive neighborhood switching is t0, and the convergence time of the neighborhood search within a single interval is t. s Average dwell time τ D The interval time is dynamically adjusted based on the urban intersections and vehicle driving conditions in the simulation experiment; ensuring: t s ≤τ D ≤t0 means that the vehicle can guarantee the continuity and stability of the planned trajectory in urban intersection scenarios.

[0064] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for variable neighborhood search and trajectory planning for autonomous driving in urban scenarios, characterized in that, Includes the following steps: S1: The vehicle is located based on the surrounding environment perception and high-precision positioning device; after the urban intersection environment is located and divided, the parking line at the urban intersection is used as the switching point for trapezoidal neighborhood search, and the parking line is used as the switching point for rectangular neighborhood search when the vehicle leaves the urban intersection scene. S2: Provide a mathematical description of the rectangular and trapezoidal neighborhoods; S3: Perform trajectory planning and optimization for the search of independent interval neighborhoods; Planning within rectangular and trapezoidal neighborhoods involves dividing and superimposing the surrounding environment within each interval of the neighborhood switching search to obtain regular shapes like rectangles and trapezoids, thereby planning a complete trajectory. The trajectory planning for independent interval neighborhood search first determines the size of a single independent interval neighborhood, and then determines the start and end points of the independent interval neighborhood based on the vehicle size. Taking the front axle center of a regular vehicle as the start point and the start point of the next independent interval neighborhood as the end point, the driving route within a single independent interval neighborhood is planned according to the fifth interpolation polynomial. S4: Use the average dwell time method to ensure the continuity and stability of the trajectory under the variable neighborhood search; The process is divided into the following steps: S41: Use the Lyapunov function to determine the stability of the trajectory within a single neighborhood and its convergence time; S42: Use the average dwell time method to ensure the stability of neighborhood search switching; To address the discontinuity during handover between different neighborhoods, an average dwell time method is introduced into the handover system to ensure the time interval between any two consecutive handovers. Greater than the average dwell time τ D This ensures the continuity and exponential stability of the switching system; the interval time Based on the dynamic adjustments made according to urban intersections and vehicle driving conditions in simulation experiments; definition N σ ( t , τ ) is 0≤ τ ≤ t The number of switches within the time period, Save(τ) D N(0) are all switches that satisfy the following inequality: Where τ D It is the average length of stay, τ. D This refers to the average dwell time of the system under a neighborhood search, where N0 is the amplitude of the switching signal; as long as the average dwell time τ D It is large enough that any number of switches within the set can guarantee the exponential stability of the switching system. When switching neighborhoods, if the search for the next neighborhood gets stuck in a local optimum, the convergence time of the neighborhood search for a single independent interval is calculated according to step S3 above. When the vehicle gives a switching signal, the average dwell time τ is guaranteed. D Greater than convergence time , that is, τ D ≥ It guarantees that when switching neighborhood searches, the next neighborhood search interval will not get trapped in a local optimum; it guarantees: ≤τ D ≤ This means that the vehicle can ensure the continuity and stability of its planned trajectory in urban intersection scenarios.

2. The method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios according to claim 1, characterized in that: In S1, locating the vehicle includes dividing the vehicle into various positions during its journey in a complex urban environment, and identifying the trigger point for the vehicle's neighborhood search conversion. The location divisions include: the straight section before entering the city intersection, the waiting area at the city intersection, the driving area in the center of the city intersection, and the exit area at the city intersection. Urban intersection waiting area: When a vehicle is about to enter an urban intersection, the area in which the vehicle can travel in a straight line from the stop line of the zebra crossing at the urban intersection is called the urban intersection waiting area. Central driving area of ​​urban intersections: The driving area from when a vehicle leaves the stop line and enters the core area of ​​an urban intersection according to traffic signals, until it leaves the stop line of the urban intersection, is called the central driving area of ​​urban intersections. Urban intersection departure zone: The driving range of a vehicle from the stop line of an urban intersection to the point of leaving the location of the urban intersection scene is called the urban intersection departure zone. After locating and dividing the urban intersection environment, the parking line leading to the urban intersection is used as the switching point for trapezoidal neighborhood search, and the parking line is also used as the switching point for rectangular neighborhood search when the vehicle leaves the urban intersection.

3. The method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios according to claim 1, characterized in that: In S2, the mathematical description of the neighborhood facilitates the division of the vehicle's perceived environment, and the complete trajectory is planned by superimposing the divisions of different neighborhoods.

4. The method for autonomous driving variable neighborhood search and trajectory planning in urban scenarios according to claim 1, characterized in that, The mathematical description of the trapezoidal search neighborhood is as follows: The mathematical expression for the trapezoidal search neighborhood is given by O(x0,y0), which represents the initial position of the intelligent vehicle. The side edges l1 and l2 of the neighborhood are both of length r, the top edge is l3, the height is s, and the bottom edge is d, as shown in the following formula: Trajectory planning within a trapezoidal neighborhood: Determine the starting and ending points within the trapezoidal search neighborhood, use quintic interpolation polynomial neighborhood planning to obtain a candidate trajectory set, and then obtain the optimal trajectory curve.

Images