改进差分进化算法在转向梯形机构优化中的应用  

Application of Improved Differential Evolution Algorithm in Steering Trapezoid Mechanism Optimization

作　　者：王琪[1] 孙慧 孙宇峰 陆凤祥 WANG Qi;SUN Hui;SUN Yufeng;LU Fengxiang(College of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang Jiangsu 2120Q3,China;Jiangsu Yueda Huanghai Tractor Manufacturing Co.,Ltd,Yancheng Jiangsu 224000,China)

机构地区：[1]江苏科技大学机械工程学院,江苏镇江212003 [2]江苏悦达黄海拖拉机制造有限公司,江苏盐城224000

出　　处：《机械设计与研究》2019年第6期173-178,共6页Machine Design And Research

基　　金：国家科技部重点研发计划资助项目(2016YFD0700900)。

摘　　要：研究了基于萤火虫变异机制的双种群差分进化算法(IGDE)对断开式转向梯形机构的优化设计。针对基本差分进化算法(DE)在寻优处理过程中经常出现收敛速度较慢甚至早熟等弊端,提出了一种基于萤火虫变异机制的双种群差分进化算法。该算法通过模仿萤火虫吸引机理,同时结合自适应柯西扰动策略求解优化最优解。通过梯形机构优化测试结果表明:与Matlab遗传算法工具箱(GA)、基本差分进化算法(DE)、随机变异差分进化算法(RMDE)和多种群自适应差分进化算法(ADEMP)相比,IGDE最优适应度值分别降低了36.08%、29.05%、17.34%和15.19%。The optimal design of the split steering trapezoid mechanism based on the two一population differential evolution algorithm(IGDE)of firefly mutation mechanism is studied.Aiming at the problems of slow convergence and even premature convergence of basic differential evolution algorithm(DE)in the process of optimization,a two-population differential evolution algorithm based on firefly mutation mechanism is proposed.The algorithm simulates the attraction mechanism of firefly and combines the adaptive Cauchy perturbation strategy to solve the optimal solution.The results of trapezoidal mechanism optimization test show that compared with the genetic algorithm toolbox(GA),basic differential evolution algorithm(DE),random mutation differential evolution algorithm(RMDE)and population adaptive differential evolution algorithm(ADEMP)of MATLAB,The optimal fitness of IGDE decreased by 36.08%,29.05%,17.34%and 15.19%respectively.

关 键 词：断开式转向梯形机构 萤火虫变异机制 自适应柯西分布 

分 类 号：U463.45[机械工程—车辆工程]