Intelligent automobile obstacle avoidance planning method

By combining the improved Golden Leopard algorithm and dynamic window algorithm, the local optimal solution and time-consuming problems in intelligent vehicle obstacle avoidance planning are solved, and more efficient and safe path planning is achieved to generate a collision-free safe obstacle avoidance path.

CN118936498APending Publication Date: 2024-11-12JIANGLING MOTORS
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410985597.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In the prior art, dynamic window algorithms have problems such as local optimal solutions and long-term obstacle avoidance planning in complex environments.

Method used

The integrated and improved Golden Leopard algorithm and dynamic window algorithm are adopted to optimize the search and development functions of the Golden Leopard algorithm, and add economic and safety and stability fitness functions to the dynamic window algorithm to carry out obstacle avoidance planning for intelligent cars.

Benefits of technology

It effectively solves the problems of local optimal solutions and time-consuming, improves the efficiency and safety of smart car obstacle avoidance planning, and generates a collision-free and safe path.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936498A_ABST
    Figure CN118936498A_ABST
Patent Text Reader

Abstract

The invention belongs to the field of automobile manufacturing, and particularly relates to an intelligent automobile obstacle avoidance planning method. Firstly, an improved golden leopard algorithm is used for initial planning, and an initial optimal global obstacle avoidance point is obtained; carrying out quadratic programming by utilizing an improved dynamic window method, sampling the driving path length, the lateral acceleration and the yaw velocity of the vehicle, and screening an optimal obstacle avoidance path through an evaluation function; and finally, a collision-free safe obstacle avoidance path is obtained. The method is accurate in planning and high in calculation precision, and has a good effect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of automobile manufacturing, and in particular relates to an intelligent automobile obstacle avoidance planning method. Background Art

[0002] Intelligent vehicle path planning aims to design a driving path for intelligent vehicles to ensure basic conditions such as driving time, collision avoidance effect and smooth and stable route from departure to destination. The goal of this planning is to provide the best path for autonomous vehicles so that they can complete driving tasks efficiently and safely. Given the complexity and diversity of the environment in which autonomous vehicles are located, including various structured and unstructured roads such as urban roads and rural roads, the planned path must be able to adapt well to this complex environment and have good collision avoidance capabilities, while improving the efficiency, safety and reliability of path planning. Path planning methods can be divided into two categories: global planning and local planning. Commonly used methods for global path planning include Dijkstra algorithm, A* algorithm, ant colony algorithm and RRT algorithm. Among them, A* algorithm and ant colony algorithm are efficient and widely used methods, but A* algorithm cannot guarantee to find the best path when the amount of calculation is too large, and the environmental adaptability of ant colony algorithm is limited. Commonly used methods for local path planning include artificial potential field method, genetic algorithm and dynamic window method. Artificial potential field method is easy to implement, fast in calculation and easy to understand, but it is easy to fall into the problem of local optimum or unable to reach the target. In actual path planning, unpredictable and complex changes need to be considered, so the planned path needs to be updated dynamically in real time to achieve good collision avoidance effects and achieve the optimal path planning goal. Summary of the invention

[0003] Aiming at the shortcomings of the dynamic window algorithm in the prior art, such as the local optimal solution and the long time-consuming obstacle avoidance planning in complex environments, the present invention proposes an intelligent automobile obstacle avoidance planning method integrating the Golden Leopard algorithm and the dynamic window. First, the improved Golden Leopard algorithm is used for initial planning to obtain the initial optimal global obstacle avoidance point; then the improved dynamic window method is used for secondary planning to sample the vehicle's driving path length, vehicle lateral acceleration and vehicle yaw angular velocity, and then the optimal obstacle avoidance path is selected through the evaluation function; finally, a collision-free and safe obstacle avoidance path is obtained. The specific technical scheme is as follows:

[0004] An intelligent vehicle obstacle avoidance planning method comprises the following steps:

[0005] Step 1: Model the vehicle driving environment;

[0006] Step 2: Through step 1, obtain the initial position of the vehicle, obstacle distribution and target point position;

[0007] Step 3: Based on the information from steps 1 and 2, use the optimized Golden Leopard algorithm to perform preliminary obstacle avoidance planning;

[0008] Step 4: Obtain the global optimal obstacle avoidance point through the initial planning of the improved Golden Leopard algorithm in step 3;

[0009] Step 5: Input the global optimal obstacle avoidance point in step 4 into the improved dynamic window algorithm to perform secondary obstacle avoidance planning;

[0010] Step 6: The vehicle generates a collision-free and safe obstacle avoidance path and reaches the target point, ending; otherwise, return to step 4.

[0011] Furthermore, the Golden Leopard algorithm is optimized in step 3 as follows:

[0012] First, the search function of the Golden Leopard algorithm is improved as shown below:

[0013]

[0014] In formula (1) and formula (2), t represents the current iteration number; It represents the position of the i-th individual (prey) in the current iteration; Respectively represent the male jackals and female jackals in the current iteration, that is, the best individual and the second best individual in the population. and They represent the relative positions of each individual i and the male jackal and female jackal in the current iteration. That is to say, there is a pair of position vectors corresponding to each individual i. E is the escape energy of the prey, rl is a random vector based on Levy distribution, and ql is the newly added Logistic chaos mapping function.

[0015] The calculation of E is as follows:

[0016] E=E1·E0·E s (3)

[0017] In formula (3), E1 is the energy of prey escape, E2 is the initial escape energy, and E s is the newly added search energy.

[0018]

[0019] In formula (4), c is a random number between [0.5, 1].

[0020] Then improve the Golden Leopard algorithm development function as shown below:

[0021]

[0022] Finally, each individual in the population updates its position function as follows:

[0023]

[0024] In formula (7), k1 and k2 represent the weighted coefficients of the optimal positions of male and female jackals.

[0025] Furthermore, the improved dynamic window algorithm in step 5 is as follows:

[0026] Based on the evaluation function of the traditional dynamic window algorithm, a new fitness function that considers economy and safety stability is added, as follows:

[0027]

[0028] In formula (8), length is the path length indicator, lat acc is the vehicle lateral acceleration index, is the yaw rate index, where α, β, and γ are the weighted coefficients of each index. By comprehensively considering the path length, vehicle lateral acceleration, and yaw rate, the safety and stability of the vehicle during driving can be greatly improved, the time consumed by the algorithm can be reduced, and the economy of the algorithm can be improved.

[0029] The present invention optimizes and improves the Golden Leopard algorithm and the dynamic window algorithm and then organically combines the two algorithms, thereby solving the problems of local optimal solutions of the traditional dynamic window algorithm and the long time-consuming obstacle avoidance planning of intelligent vehicles in complex environments, and using it in the obstacle avoidance planning of intelligent vehicles has a good effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A schematic diagram of the flow chart of the fusion algorithm method of the present invention;

[0031] Figure 2 Comparison diagram of obstacle planning paths between the fusion algorithm of the present invention and the traditional algorithm in a complex obstacle environment;

[0032] Reference numerals: Figure 2 middle As the starting point, For the end point. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Example

[0035] An intelligent vehicle obstacle avoidance planning method comprises the following steps:

[0036] Step 1: Model the vehicle driving environment;

[0037] Step 2: Through step 1, obtain the initial position of the vehicle, obstacle distribution and target point position;

[0038] Step 3: Based on the information from steps 1 and 2, use the optimized Golden Leopard algorithm to perform preliminary obstacle avoidance planning;

[0039] Step 4: Obtain the global optimal obstacle avoidance point through the initial planning of the improved Golden Leopard algorithm in step 3;

[0040] Step 5: Input the global optimal obstacle avoidance point in step 4 into the improved dynamic window algorithm to perform secondary obstacle avoidance planning;

[0041] Step 6: The vehicle generates a collision-free and safe obstacle avoidance path and reaches the target point, ending; otherwise, return to step 4.

[0042] Furthermore, the Golden Leopard algorithm is optimized in step 3 as follows:

[0043] First, the search function of the Golden Leopard algorithm is improved as shown below:

[0044]

[0045] In formula (1) and formula (2), t represents the current iteration number; It represents the position of the i-th individual (prey) in the current iteration; Respectively represent the male jackals and female jackals in the current iteration, that is, the best individual and the second best individual in the population. and They represent the relative positions of each individual i and the male jackal and female jackal in the current iteration. That is to say, there is a pair of position vectors corresponding to each individual i. E is the escape energy of the prey, rl is a random vector based on Levy distribution, and ql is the newly added Logistic chaos mapping function.

[0046] The calculation of E is as follows:

[0047] E=E1·E0·E s (3)

[0048] In formula (3), E1 is the energy of prey escape, E2 is the initial escape energy, and E s is the newly added search energy.

[0049]

[0050] In formula (4), c is a random number between [0.5, 1].

[0051] Then improve the Golden Leopard algorithm development function as shown below:

[0052]

[0053] Finally, each individual in the population updates its position function as follows:

[0054]

[0055] In formula (7), k1 and k2 represent the weighted coefficients of the optimal positions of male and female jackals.

[0056] Furthermore, the improved dynamic window algorithm in step 5 is as follows:

[0057] Based on the evaluation function of the traditional dynamic window algorithm, a new fitness function that considers economy and safety stability is added, as follows:

[0058]

[0059] In formula (8), length is the path length indicator, lat acc is the vehicle lateral acceleration index, is the yaw rate index, where α, β, and γ are the weighted coefficients of each index. By comprehensively considering the path length, vehicle lateral acceleration, and yaw rate, the safety and stability of the vehicle during driving can be greatly improved, the time consumed by the algorithm can be reduced, and the economy of the algorithm can be improved.

[0060] The specific process is as follows Figure 1 shown.

[0061] Comparative Example

[0062] Using the traditional algorithm, that is, the traditional dynamic window algorithm, the proposed fusion algorithm is compared with the traditional algorithm in the planning path of the complex obstacle environment. Figure 2 As shown, the black area is the obstacle.

[0063] Table 1 is a comparison of three economic indicators of the obstacle avoidance paths planned by the fusion algorithm of the present invention and the traditional algorithm.

[0064] Table 1

[0065]

[0066] From the above comparison, it can be seen that the fusion algorithm of the present invention is more accurate in planning obstacle avoidance paths than the traditional algorithm, and it takes less time and runs less times, thus having a good effect.

[0067] The above describes in detail the preferred implementation of this patent, but this patent is not limited to the above implementation. Various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of this patent.

Claims

1. An intelligent vehicle obstacle avoidance planning method, Features: The steps include: Step 1: Model the vehicle driving environment; Step 2: Through step 1, obtain the initial position of the vehicle, obstacle distribution and target point position; Step 3: Based on the information from steps 1 and 2, use the optimized Golden Leopard algorithm to perform preliminary obstacle avoidance planning; Step 4: Obtain the global optimal obstacle avoidance point through the initial planning of the improved Golden Leopard algorithm in step 3; Step 5: Input the global optimal obstacle avoidance point in step 4 into the improved dynamic window algorithm to perform secondary obstacle avoidance planning; Step 6: The vehicle generates a collision-free and safe obstacle avoidance path, reaches the target point, and ends; Otherwise, return to step 4.

2. According to the intelligent vehicle obstacle avoidance planning method of claim 1, Features: The optimized Golden Leopard algorithm in step 3 is as follows: First, the search function of the Golden Leopard algorithm is improved as shown below: In formula (1) and formula (2), t represents the current iteration number; It represents the position of the i-th individual (prey) in the current iteration; They represent the male jackal and female jackal in the current iteration, i.e., the optimal individual and suboptimal individual in the population; E is the escape energy of the prey; rl is a random vector based on Levy distribution; ql is the newly added Logistic chaotic mapping function; The calculation of E is as follows: E=E 1 ·AND 0 ·AND s (3) In formula (3), E 1 Energy for prey to escape, E 1 is the initial escape energy, Es is the newly added search energy; in Formula (4) c is a random number between [0.5, 1]; Then improve the Golden Leopard algorithm development function as shown below: Finally, each individual in the population updates its position function as follows: In formula (7), k 1 and k 2 represents the weighted coefficient of the optimal position of male and female jackals.

3. According to the intelligent vehicle obstacle avoidance planning method of claim 1, Features: The improved dynamic window algorithm in step 5 is as follows: Based on the evaluation function of the traditional dynamic window algorithm, a new fitness function that considers economy and safety stability is added, as follows: In formula (8), length is the path length indicator, lat acc is the vehicle lateral acceleration index, is the yaw rate index, where α, β, γ are the weighting coefficients of each index.