Research Of Path Planning For Mobile Robots Based On Improved Ant Colony Algorithm

Path planning of mobile robots is one of key issues in robotics research. It is always afocus topic for researchers all over the world. However, traditional methods of path planningall have their own drawbacks respectively. To find a superior algorithm for the problembecomes a study hotspot in the field. The ant colony algorithms(ACO) is a new bionicsoptimization algorithm developed in the past decade,it shows excellent performance andgreat potential for development when solving many complex problems．This thesis presentsan ant colony algorithm for solving the path planning problem in static mobile robotenvironments.Firstly, this thesis establishes the environment model by using grid method and makespath planning in grid-based environment by basic ACO. Secondly, three improved strategiesare proposed to overcome certain defects: (1) To eliminate stagnation behavior of executingalgorithm, a full convex processing strategy is proposed. The proposed strategy can make antcolony search the path incessantly and ensure reliability and robustness of the algorithm. (2)When two ants with the different searching direction meet, a meeting strategy is used to get anew path by using the records information in forbidden tables of the two meeting ants. Thenthe new method can get enough cooperation of ant colony and the speed of getting new pathcan be improved. (3) To eliminate misleading of the pheromone that ants release in initialstage of path planning, ACO with punishment strategy is proposed. Finally, a new algorithmis proposed based on the improved ACO and genetic algorithm (GA): ACO-GA algorithm,the new algorithm can be introduced to path planning of mobile robots.In order to verify the effectiveness of all proposed algorithms, a path planning simulationsystem for mobile robots is designed and implemented based on VC++ 6.0 developmentenvironment. Simulation results show that the proposed algorithm can reach a preferable performance.