| 周荣敏,雷延峰.基于改进单亲遗传算法的树状管网布置优化[J].水利学报,2012,43(10): |
| 基于改进单亲遗传算法的树状管网布置优化 |
| Optimal layout of tree pipe networks based on improved single parent genetic algorithm |
| |
| DOI： |
| 中文关键词: 改进单亲遗传算法 树状管网 优化布置 |
| 英文关键词: improved single parent genetic algorithm tree pipe network optimal layout |
| 基金项目: |
|
| 摘要点击次数: 2182 |
| 全文下载次数: 406 |
| 中文摘要: |
| 改进单亲遗传算法采用Kruskal算法和Dijkstra算法进行群体初始化代替随机群体初始化过程；采用赌轮盘选择和单亲换位算子作为主要遗传算子，取消选择率、换位率和单亲逆转算子， 使算法结构更加简洁明了；增设单一化的最优群体，并自动更新最优群体适应度值的下限。研究表明，通过一系列改进，在同样能获得最优解的前提下，程序运行时间由70s 缩短到5s，最大遗传代数由500 代以上缩短到100 代以下，改进单亲遗传算法（ISPGA）的性能提高显著。 |
| 英文摘要: |
| The improved single parent genetic algorithm adopts Kruskal algorithm and Dijkstra algorithm to initialize the populations instead of the random population initializing process， and adopts the roulette wheel and the single parent transposition operator as the main genetic operators，cancels the selection rate， the transposition rate and the single parent reverse operator to make the structure of the algorithm simpler and clearer. It sets the optimum populations and automatically updates the lowest limit value of minimum fitness of the optimum populations. The results show that on the premise of achieving the optimum solution，it reduces the program run time from 70 seconds to 5 seconds， reduces the maximum generation from above 500 generations to below 100 generations through a series of improvements， the efficiency of ISPGA is enhanced remarkably. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |