In order to resolve the problems that the standard PSO algorithm is apt to be easily trapped in local optima and the LDW-PSO algorithm cannot adapt to the complex and nonlinear optimization,the paper proposes a modified particle swarm optimization algorithm based on the information entropy theory,named EPSO.The information entropy value is used by EPSO to determine the inertia weights,which make the algorithm have the ability of "explore" and "exploit" adaptively.The new algorithm is realized for the parameter selection of support vector machine.The simulation results prove that the proposed EPSO is stable.Compared with PSO and LDW-PSO,EPSO enhances the ability of escaping from local optimal solution,and becomes more feasible in engineering application.%针对标准PSO算法易陷入局部最优化和LDW-PSO算法不能适应复杂、非线性优化的问题,提出了一种基于信息熵理论的改进粒子群算法(EPSO).该方法利用信息熵值确定惯性权值,使之具有自适应地调整“探索”和“开发”的能力.将新算法应用于调制模式识别中SVM分类器最优参数值的确定,仿真研究实明,该算法性能稳定.与标准PSO和LDW-PSO算法相比,EPSO算法有效增强了跳出局部最优解的能力,具有较好的工程应用性.
展开▼
