【摘要】 研究三维空间机器人路径规划问题,由于系统求解时间较长、过早失去解的多样性、易陷入局部最优、个体适应度较差等问题,通过构建三维工作空间模型、引入变异算子和搜索无碰路径策略来解决,提出适宜于三维机器人路径规划的一种变异算子蚁群算法(Mutation Operator Ant Colony Algorithm,MOACA)。MOACA是一种关于模型构造的启发式搜索算法,算法在改进启发式函数设计、选择概率确定、信息素更新策略等基础上,引入逆转变异和插入变异算子,通过选择逆转点反序排列部分路径节点和随机插入路径节点的方法搜索无碰路径,对蚁群算法进行了局部优化改良。仿真结果表明,MOACA在搜索路径、收敛时间、适应度等方面较传统蚁群算法有明显改善,算法是有效可行的。更多还原

【Abstract】 In view of the robot path planning in 3D space,there are so many problems,such as long computation,early loss of the solution diversity,easy to fall into local optimum,and poor individual fitness. To solve these problems,the Mutation Operator Ant Colony Algorithm( MOACA) is a good solution. The MOACA is made up of constructing 3D workspace model,introducing mutation operator and the strategy of searching non- collision path.Based on the construct of the model,MOACA is a heuristic search algorithm. It introduces reversal mutation and insertion mutation operator on the basis of the improved heuristic function design,selects probability and pheromone updating strategy. Through selecting turning point to array the nodes in reverse order and inserting random nodes to find the non- collision path,the optimization of ant colony algorithm is partly improved. The simulations show that the MOACA not only has good effectiveness and feasibility,but also is improved in the areas of path length,convergence speed and fitness.更多还原