Research Of Path Planning Based On Improved RRT Algorithm

The field of robotics,especially mobile robotics,has been developed significantly in recent years,and path planning is one of the key technologies for mobile robots.Among them,Rapidly-exploring Random Trees(RRT)algorithm has the advantages of fast search speed and no pre-processing of the environment.However,it has the problems of difficult sampling and slow convergence in dealing with the environment where narrow channels exist,so two improved algorithms,Mix-RRT algorithm and ED-RRT algorithm are proposed.The Mix-RRT algorithm focuses on solving the problem that the RRT algorithm cannot identify the entrances of narrow channels.Adding a target bias strategy to guide the random tree to expand in favor of the target point;adding an entrance identification strategy to guide the random tree to identify the entrance of the narrow channel and make it enter the narrow channel smoothly;adding a range restriction strategy to control the selection of sampling points improves the planning efficiency of the improved algorithm.The ED-RRT algorithm is based on Mix-RRT algorithm with three improvement strategies.Among them,the forward detection strategy optimizes the problem that the MixRRT algorithm cannot distinguish between entrances and traps;the dynamic circle strategy optimizes the problem that it is difficult to sample in narrow channels;the variable target point strategy optimizes the problem that it is difficult to expand in tortuous channels by changing the temporary target point to guide the random tree through the tortuous narrow channels.After the initial path is planned,a greedy algorithm is used to reduce the redundant nodes to make the planned path smoother.Finally,the simulation experiments comparing the two improved algorithms with the RRT algorithm and Bias-RRT algorithm using MATLAB show that the success rates of the improved algorithms all remain above 90%,and the time consumed for planning,the number of iterations,and the path length are better than those of the control group,verifying that the improved algorithms are effective and feasible for path planning in such environments.Figure 39;Table 8;Reference 63...