基于改进型蚁群算法的AUV路径规划

摘要

Path planning is meant by looking for a path without collision from the start point to target point in an environment which the obstacle is known.Improved ant colony algorithm is extended to the area of path planning for autonomous underwater vehicle,namely AUV for short.To improve the drawbacks of traditional ant colony algorithm in practical application,an improved ant colony algorithm based on reinforcement learning is proposed.The ant algorithm's shortcomings of slow in search and easy in trapping into local optimal solution are improved with the usage of reward and punishment in the update of ant colony pheromone.Improve the search speed and optimization ability of the algorithm can obviously improve the efficiency of path planning.Simulation results verify the effectiveness of the improved algorithm.%在已知障碍物的环境中寻找一条从起点到终点的无碰路径即为路径规划.扩展改进型蚁群算法应用背景,应用于智能水下机器人(AUV)的路径规划.为改善传统蚁群算法在实际应用中的不足,提出加入再励学习机制的改进型蚁群算法.通过对蚁群信息素更新实行奖惩制度后,改善蚁群算法搜索求解缓慢,易陷入局部最优解而产生停滞现象,提高算法的搜索速度及寻优能力,能够明显地提高路径规划的效率.仿真结果验证了改进算法的有效性.