Path Planning Of Intelligent Vehicles On Uneven Terrain

Since the advent of the mobile robot in 1860 s,it has been widely studied by researchers from many countries. With the application of computer and the development of sensor technology,a large number of world famous companies began to develop mobile robot platform. Mobile robots gradually transition from the laboratory to social life,applied to agricultural production,marine development,social services and even life entertainment and other fields. Along with the progress of technology and the increasing demand of reducing casualties, the application of unmanned vehicle in the military field has gradually got the attentions of the countries,and all increased the research strength of the unmanned combat platform. As early as the Second World War,the military remote control ground vehicles have come out, and take on tasks of mine and explosive removal,getting rid of obstacles,opening circuit and other high-risk tasks. After 1990 s,the autonomous vehicle technology and other key technologies have a breakthrough progress,resulting in the rapid progress of unmanned ground vehicles. The United States,Russia,Britain,France and other countries have carried out different levels of unmanned vehicle research and development programs.Different from the mobile robot running in the urban environment,the operating environment of the unmanned combat platform is mostly unstructured terrain. These environments are complex and diverse,including steep slopes,pits,water and other obstacles. which exerts on the operation of unmanned vehicles a huge challenge. In the first place,it is a basic problem that how to correctly identify the hazardous area in the environment. Secondly, on the basis of the existing environmental model, it is a question of how to properly plan a safe route for the vehicle to follow, and meeting the real-time requirements. In the existing research on the path planning of unmanned vehicles,most of them are based on the plane environment, and most of them take the shortest path length or the shortest time as the optimal criteria. In the off-road environment, the complex terrain conditions bring more factors to the path planning,and continuing to use the simple path optimization criteria of the plane environment will not be enough to obtain satisfying results.In this paper,the application of intelligent vehicles in off-road environment is analyzed,and the methods of environment mapping are briefly introduced. On the basis of the accurate acquisition of the environment model,this paper focused on the methods of path planning. In the environment modeling,the terrain features of various types of road barriers,such as steep slopes,pits,water bodies and so on,are used to build up the grid map of the environment. On the basis of the grid map, the combination of global planning and local planning is used to search the path. In global planning,to deal with the problem that large environment makes the search algorithm converge slowly,the Voronoi diagram is used to extract the global roadmap which can greatly reduce the search space of the algorithm, at the same time to solve the potential optimality invalidation problem of roadmap search, the Voronoi roughness map which gives full consideration to the overall area conditions surrounding the roadmap is introduced, so as to ensuring that the algorithm is able to find the really optimal global path for the vehicle. Local planning tracks global search results,to ensure the local planning can maintain greater flexibility when following the global path, the safe corridor is extracted based on the global path. In order to ensure the executability of local planning results, a feasible path unit set is generated base on the bicycle model to form the path. To further improve the rapid response capability of the local planning, the dynamic window approach is used to reduce the environment area under consideration, which greatly improves planning efficiency of the algorithm. Through the global and local planning strategy,the vehicle can avoid the dangerous area in the environment and at the same time effectively carry out the local tracking process, which realizing the safe and efficient running of intelligent vehicles in the uneven terrain.Through the simulation experiments, it is proved that the method can quickly generate a global safety path in the uneven environment,at the same time,local planning can generate the correct feasible path and steer the vehicle to precisely track it,thereby ensuring that the vehicle can safely and accurately reach the target, thus well solves the problem of path planning in uneven environment for intelligent vehicles.