基于几何规则的异类蚁群优化算法

Heterogeneous feature ant colony optimization algorithm based on geometric rules

针对复杂环境下自动导引小车路径规划存在收敛速度慢、极易陷入局部最优的缺点,提出一种基于几何规则的异类蚁群优化(GR-HFACO)算法。首先,为加快算法收敛速度,利用几何规则非均匀分配初始信息素,设置双向并行搜索机制;其次,引入具有观点采择能力的蚂蚁高效协同工作,改善路径全局的随机搜索特性;最后,为平衡算法的收敛性及全局性,在更新环节引入信息素负反馈环节以及交叉操作,并证明了GR-HFACO算法具有全局收敛性。仿真结果表明,该算法的收敛速度以及全局搜索性能显著优于目前流行的ACON、TWPSS-ACO、SoSACO-v2、Sci-ACO和HHACO算法。

For path planning of automated guided vehicle in the complex environment , this paper proposed a heterogeneous feature ant colony optimization algorithm based on geometric rules to solve the problem of low efficiency.Firstly,to accelerate convergence,this paper presented an inhomogeneous distribution of initial pheromone based on geometry rules.And then this paper introduced a method of incorporating perspective-taking ability to generate differently acting ant colonies in order to increase search diversity. Besides, for maintaining rapidity and randomness, this paper adopted an update rule of pheromone negative feedback and cross operator. Finally, this paper proved the convergence of GR-HFACO algorithm. Simulation results show that the path planning efficiency of the proposed algorithm is outperform those of popular ACON, TWPSS-ACO, SoSACO-v2, Sci-ACO and HHACO algorithms.