| Along with the continuous progress and development of mobile robotics,mobile robots are also used in the field environment such as battlefield,mountainous area,and even outer planet,and the demand of society for intelligent robots is increasing.In the application of mobile robots,path planning is a very critical issue.Path planning is the basic element of autonomous behavior of mobile robots,and in this paper,we study the path planning method for mobile robots in the field environment.Firstly,the environment modeling methods commonly used for path planning are introduced.For the characteristics of the field environment,digital elevation models conforming to the characteristics of the natural environment are generated by the methods of terrain simulation and mountain peak simulation,respectively.Secondly,based on the analysis of the basic working principle of RRT algorithm,the node expansion rules of RRT algorithm are improved for the special modeling way of field environment and the planning demand of flat path of mobile robot.A new node acceptance criterion according to probability acceptance is proposed,and pitch and turn angle constraints are set for the mobile robot,and a method of path smoothing through Bessel curves is designed.Simulation tests are conducted in different mountainous environments,and the simulation test results show that the improved RRT algorithm can plan a smooth path with high planning efficiency in the field environment.Thirdly,research is conducted on dynamic path planning for field environments with sudden threats.The gray wolf algorithm is selected for global path planning to obtain the initial path,and when the mobile robot encounters a sudden moving obstacle in the course of driving,the artificial potential field method is activated for local obstacle avoidance planning,and a hybrid path planning algorithm with global and local fusion is designed.For the global path planning part of the hybrid algorithm,this paper proposes chaotic initialization of the gray wolf population and the introduction of a differential evolution strategy in the gray wolf population to further improve the global search capability of the algorithm and accelerate the convergence speed.The simulation test results show that the hybrid path planning of the improved gray wolf algorithm and the artificial potential field method has obvious obstacle avoidance effect against the sudden threats and is suitable for the mobile robot to realize dynamic path planning in the field environment.Finally,a path planning algorithm suitable for collaborative multi-mobile robots in a field environment is proposed to implement this process using an improved firefly algorithm.In order to further improve the solution accuracy and stability of the firefly algorithm,an adaptive step size strategy is proposed.Simulation test results show that this improved algorithm can successfully plan the respective paths for multiple mobile robots and meet the dynamic obstacle avoidance requirements,and it has higher convergence accuracy and planning speed compared with algorithms such as particle swarm algorithm. |
