文献类型：期刊文献

英文题名：Physarum-energy optimization algorithm

作者：Feng, Xiang[1,2];Liu, Yang[1];Yu, Huiqun[1];Luo, Fei[1]

机构：[1]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China;[2]Shanghai Jiao Tong Univ, Smart City Collaborat Innovat Ctr, Shanghai, Peoples R China

年份：2019

卷号：23

期号：3

起止页码：871

外文期刊名：SOFT COMPUTING

收录：;EI(收录号：20173604132136);WOS:【SCI-EXPANDED(收录号:WOS:000457326400011)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant Nos. 61472139 and 61462073, the Information Development Special Funds of Shanghai Economic and Information Commission under Grant No. 201602008, the Open Funds of Shanghai Smart City Collaborative Innovation Center.

语种：英文

外文关键词：Physarum optimization algorithm; Energy mechanism; Age factor; Traveling salesman problem

摘要：In general, the existing evolutionary algorithms are prone to premature convergence and slow convergence in coping with combinatorial optimization problems. So an intelligent optimization algorithm called physarum-energy optimization algorithm (PEO) is proposed and put TSP as the carrier in this paper. This algorithm consists of four parts: the physarum biological model, the energy model, the age factor model and the stochastic disturbance model. First, the high parallelism of PEO is enlightened from the physarum's low complexity and high parallelism. Second, we present an energy mechanism model in PEO, which is mainly to develop the shortcomings of existing algorithm, such as slow convergence and lack of interaction capability. Third, inspired by the characteristic of ants' spatiotemporal variations, the age factor mechanism is introduced to raise search capacity, which can control the convergence speed and precision ability of PEO. In addition, in order to avoid premature convergence, the stochastic disturbance mechanism is adopted into PEO. And also the feasibility and convergence of PEO has been analyzed and verified theoretically. Moreover, we compare the algorithm and other algorithms to TSPs of diverse scope. The experiment results show that PEO has the advantages of excellent global optimization, high optimization accuracy and high parallelism and is significantly better than other algorithms.