Research On The Safe Path Planning Of Ground Mobile Robot

Path planning is the key part of Ground Mobile Robots’ control system and it has been widely used in intelligent sweeping robots, automated carrier robot in automatic logistics center and robots on other planets. Path planning calculates the optimal path for mobile robot to move to destination and finish the mission. It decides whether a mobile robot can complete its task and the safety of a mobile robot during the process of moving to the destination.Now the existing D*Lite path planning algorithm has the advantages of using simply and easy to understand. But the existing D*Lite algorithm’s results have unsafe path for mobile robot’s moving and the results also need to be further optimized. By learning the widely used A*, LPA* and D*Lite algorithm’s principles, this paper concludes safety improvement and path optimization methods. Based on movement characteristics of Ground Mobile Robot, this paper optimizes existing D*Lite algorithm and applied the simulation results to Traveler II robot.First, introduce three algorithms that are widely used in mobile robot’s path planning:A*, LPA* and D*Lite. Learn three algorithms’path planning method and the use of heuristic search and incremental search. Based on comparisons of simulation results and the advantages and disadvantages of the three algorithms, summarize D*Lite is more suitable for Ground Mobile Robot’s path planning in dynamic environment since its search direction is from current position to the target position.Secondly, introduce the security problems of existing D*Lite algorithm:traverse across obstacle’s sharp corner and traverse in between two obstacles. Give methods to detect and improve the unsafe paths. And then make temporary marks for complex obstacles to reduce the numbers of search grids in dynamic environment. Using reversing fusion method to optimize existing D*Lite algorithm’s result. The optimized D*Lite can reduce the length of path and abolish limitation of turning angle which is limited to multiple of 45 degree in existing D*Lite. The optimized D*Lite algorithm has a smooth path planning result which is more suitable for robots’ path planning. Simulation results show the correctness of optimized D*Lite.Finally, program D*Lite algorithm in Visual C++ development platform and complete the path planning on the existing lab’s map. Then, adding new obstacles in lab’s map to verify optimized D*Lite algorithm’s stability in complex environment. Simulation results show optimized D*Lite still can plan the path in complex environment, at the same time the path is safer, shorter and smoother. Results of optimized D*Lite prove the correctness of safety improvement and path optimization. At last, apply the path planning results to Traveler Ⅱ Ground Mobile Robot. Traveler Ⅱ moves according to the path planning result of optimized D*Lite and verify the improvements of optimized D*Lite algorithm in lab’s environment.