中文摘要：

提出了一种复杂静态环境下的移动机器人避碰路径规划的改进蚁群算法，基于栅格法的工作空间模型，模拟蚂蚁的觅食行为；针对路径规划的需要，搜索过程采用了蚂蚁回退策略、目标吸引策略、参数自适应调整和路径优化策略；利用蚂蚁回退策略和惩罚函数使得蚂蚁能够顺利跳出陷阱，并且在下一次搜索中不再选择此路径，从而避免了遇到陷阱时形成的路径死锁情况，同时也提高了最优路径的搜索效率；仿真试验结果表明，该算法能迅速规划出最优路径。

英文摘要：

An improved ant colony algorithm was proposed to plan an optimal collision--free path for mobile robot in complicated static environment. Based on the workspace model with grid method, the foraging behavior of ant colony was simulated. Furthermore, the strate- gies of backspace from traps, goal attraction, adjusting parameters adaptively and path optimization were applied to path planning of mobile robot. The strategy of backspace from traps and punishment function enabled ant jump out of traps successfully, and made the ant don＇t choose this path in next search, so it avoided path--locked situation as well as improved the efficiency of planning optimal path. The simula- tion results showed that the best path could be rapidly found.