Trajectory Planning Algorithm For Mining Trucks Based On Improved A* Algorithm

In recent years,with the rapid development of the country’s economic and technological level,the technological and intelligent development of the industry has received more and more attention from all sectors of society.One of the most important directions is autonomous driving,because the machine executes with certainty,so the implementation of unmanned vehicles can greatly solve the problem of driver manpower and the safety problems caused by human reasons.However,there is still a long way to go before the autonomous driving is fully implemented,and there are great challenges in the direction of unmanned single-vehicle.With the progress and development of science and technology,intelligence has slowly penetrated into traditional industries.The mining industry has always been an important strategic direction for our country,but the traditional mining industry has many problems such as high danger,high work intensity and worse working environment,etc.Due to the above problems,mining enterprises currently have difficulties in recruiting workers,high costs and high pollution,so how to achieve unmanned mines is the key to solve the painful problems of the traditional mining industry.And in which autonomous driving is an important part of solving the mine transportation problem.Path planning,as a core part of the autonomous driving framework,also faces complex challenges.In this paper,we propose a path planning algorithm for complex mining conditions,for which there were few mining-specific planning algorithms in the industry before.The algorithm has been tested offline with good results and has been deployed on several mining dump trucks with passing by stable obstacle during real-world testing.This paper first briefly describes the framework structure of the autonomous driving,and then introduces the intelligent modification of the mining dump truck,which has been modified to achieve initial perception and localization of the environment in a complex mining environment,providing the basis for the path planning algorithm.Then an improved path planning algorithm based on the traditional A*algorithm is proposed for the problems of safety and real-time in the local path planning of unmanned vehicles in intelligence mines.This paper firstly changes the traditional A* algorithm path point constraint from a two-dimensional(x,y)to three-dimensional(x,y,yaw),being the generated trajectory connected by a curve instead of a straight line.Secondly,the heuristic function of the algorithm is optimised due to the change in the way the algorithm is extended.The heuristic function consists of a distance cost function,a curvature cost function and a heading angle change rate cost function.Thirdly,the initial path is optimised in order to improve the smoothness and safety of the trajectory.In this paper,innovative optimisation methods are proposed for the special working conditions of deep pits that exist in mining,which can effectively improve productivity and reduce energy consumption.The algorithm is first tested on a large scale in a simulation environment to test the completeness,safety and real-time performance of the algorithm.The algorithm is then put into use in a real-world scenario,with good results,and can achieve the function of bypassing obstacles in mining areas with autonomous driving.Compared with the traditional Dijkstra algorithm and the widely used Lattice algorithm,the algorithm proposed in this paper is significantly better in terms of planning speed and path quality,and is more suitable for complex environments in mining areas.