应用改进蚁群算法的同心分布喷丝孔检测路径规划  

Route planning for concentric spinneret inspection based on improved ant colony algorithm

作　　者：郭隽侠 陶建峰[1] 刘成良[1] 

机构地区：[1]上海交通大学机械与动力工程学院,上海200240

出　　处：《纺织学报》2017年第11期124-130,136,共8页Journal of Textile Research

摘　　要：为提高微孔同心分布喷丝板的检测效率,提出一种基于改进蚁群算法的微孔检测路径规划方法。该方法针对传统蚁群算法收敛速度慢、易陷入局部最优等缺陷进行优化改进:重新定义微孔间的距离以适应典型喷丝板检测仪运动特点;采用最近邻法设定初始信息素浓度表,使蚁群算法在相同的迭代次数等参数下求得更优的路径结果;通过路径尖端去除处理对蚁群算法的结果进一步优化,得到了优化的微孔检测路径。为验证算法的有效性,以典型的微孔同心分布喷丝板为算例进行检测路径的规划计算。结果表明,所提出的算法具有较快的收敛速度,优化所得路径与传统逐圈检测路径相比缩短路径长度约18%,可显著提高对应喷丝板的检测效率。In order to improve the inspection efficiency of concentric-distribution spinneret, an inspection route planning method based on improved ant colony algorithm was proposed. In order to overcome the shortcomings of conventional algorithm including low convergence speed and sinking into local optimum,this method redefined the distance between micropores to meet the characteristic of typical spinneret inspectors,set up the initial pheromone concentration table by the nearest neighbor method to obtain better results with other parameters unchanged,and further optimized the results by path peaks removal process. To verify the effectiveness of the proposed algorithm,calculation of the route planning for a typical concentric-distribution spinneret was carried out. The results show that the proposed algorithm has higher convergence speed,and it can shorten the route length by about 18% compared with the conventional inspection route and improve the inspection efficiency of matching spinneret.

关 键 词：喷丝板检测 蚁群算法 信息素浓度 最近邻法 最优化 

分 类 号：TS173.8[轻工技术与工程—纺织材料与纺织品设计] TP391[轻工技术与工程—纺织科学与工程]