基于改进人工蜂群算法的船舶管路路径寻优算法分析  

Analysis of ship pipeline routing optimization algorithm based on improved artificial bee colony algorithm

作　　者：李铁骊[1] 王文双 刘海洋 杨远松 林焰[1] LI Tieli;WANG Wenshuang;LIU Haiyang;YANG Yuansong;LIN Yan(School of Naval Architecture Engineering,Dalian University of Technology,Dalian 116024,China;CNNC Key Laboratory of Green Construction Technology and Equipment,Beijing 101300,China;China Nuclear Industry 23 Construction Co.,Ltd,Beijing 101300,China)

机构地区：[1]大连理工大学船舶工程学院,辽宁大连116024 [2]中核绿色建造技术与装备重点实验室,北京101300 [3]中国核工业二三建设有限公司,北京101300

出　　处：《中国舰船研究》2024年第2期1-12,共12页Chinese Journal of Ship Research

基　　金：中核绿色建造技术与装备重点实验室开放基金(CNNC-STGCL-KFKT-2022-001)。

摘　　要：[目的]人工蜂群(ABC)算法具有控制参数少、局部寻优能力强、收敛速度快的特点,但在解决路径寻优问题方面,存在容易陷入局部最优的缺陷。为解决船舶管路系统中的管路路径规划问题,提出一种改进的人工蜂群(IABC)算法。[方法]在传统人工蜂群算法的基础上,在跟随蜂的更新机制中引入遗传算子中的交叉操作,并对交叉算子的交叉概率采用自适应的策略;通过对种群进行的交叉操作寻找全局范围内的新解,并改进侦察蜂寻找新路径的方式,由原来的对路径经过的点进行更新改为对路径中的“路段”进行更新;随后,提出一种适应于解决分支管路路径寻优的改进人工蜂群协同进化算法。[结果]实例验证表明,改进后的人工蜂群算法相比标准人工蜂群算法其路径布置效果能够提升32.3%~37.4%,收敛速度能够提升17.7%~29.9%。[结论]无论是解决单管路还是分支管路,改进后的人工蜂群算法相比传统的人工蜂群算法求解质量更高、收敛速度更快、稳定性更好。[Objective]The artificial bee colony(ABC)algorithm has such characteristics as few control parameters,strong local optimization ability and fast convergence speed.However,when solving path optimization problems,it can easily fall into local optimal solutions.In order to solve the problem of pipeline routing in a ship pipeline system,an improved artificial bee colony(IABC)algorithm is proposed.[Method]Based on the traditional artificial bee colony algorithm,the crossover operation of genetic operators is introduced into the update mechanism of following bees,and an adaptive strategy is adopted for the crossover probability of the crossover operator.The crossover operation on the population is used to find new solutions in the global range.The way scout bees search for new paths is improved from updating the points that the path passes to updating the"road sections"in the path.This paper proposes an artificial bee colony co-evolution algorithm for solving the optimization of branch pipeline paths.[Results]Compared with the standard artificial bee colony algorithm,the improved algorithm can improve the path layout effect by 32.3%-37.4%and the convergence speed by 17.7%-29.9%.[Conclusion]The improved artificial bee colony algorithm proposed herein has higher solution quality,faster convergence speed and better stability than the traditional artificial bee colony algorithm for a single pipe or branch pipe.

关 键 词：船舶管路 人工蜂群算法 路径规划 协同进化 

分 类 号：U664.84[交通运输工程—船舶及航道工程]