Optimal synthesis of mechanisms using repellency evolutionary algorithm

摘要

Traditional metaheuristics are easy to fall into local optimum when solving complex dimensional synthesis problems. To solve this problem, a repellency evolutionary algorithm (REA) is proposed in this paper. The REA includes two repulsive mutation behaviors. The first one is that two parent individuals are considered as repulsive sources, and the offspring individual is repelled by parent. The other one is that one parent individual is considered as the source of repellency and repels the other parent individual, then the offspring learn their behaviors. These two mutation behaviors have the following common characteristics. First, the population no longer learns from the current global optimum to avoid population falling into the local optimum region. Second, the offspring search in any direction except the location of the parent. To demonstrate the efficiency of REA, CEC2014 and five case studies of dimensional synthesis were conducted. Seven novel metaheuristics were implemented to compare with REA. Moreover, their solutions were compared with that obtained by other algorithms from previous literature. The experimental results validated the practicability and performance advantage of REA in solving benchmarks and dimensional synthesis problems.