摘要
DE算法是一类基于种群的启发式全局搜索技术,该算法原理简单,控制参数少,鲁棒性强,具有良好的优化性能。利用差分进化算法对Wiener模型参数进行辨识,把辨识问题等价为以估计参数为优化变量的非线性极小值优化问题,并分析了算法中种群规模NP、缩放因子F、交叉概率CR等控制参数对辨识过程中的全局并行搜索能力和收敛速度的影响,以保证算法的全局收敛性。对Wiener模型的数值仿真结果表明了DE算法在参数辨识问题中的有效性,以及较PSO算法更强的非线性系统辨识能力。
Differential evolution (DE) is a heuristic global optimization technique based on population. The algorithm is simple, re- quires few control parameters, is strong robust, and has good optimization performance. DE was used to identify parameters of Wiener model. The identification problem was equivalent to nonlinear minimization problem with the estimated parameters as the optimization variables, and analyzed the influence of global parallel search ability and convergence speed as to the population size NP, scaling factor F and crossover factor CR of important control parameters in the process of identification, ensured the global convergence. A numerical simulation results of Wiener model show that DE is effective in parameter identification problems and stronger identification ability than PSO in nonlinear system.
出处
《控制工程》
CSCD
北大核心
2012年第5期900-904,共5页
Control Engineering of China
基金
国家自然科学基金项目(21206053)
中国博士后基金资助项目(2012M511678)
江苏高校优势学科建设工程资助项目(PAPD)
关键词
参数辨识
WIENER模型
差分进化算法
粒子群算法
parameter identification
Wiener model
differential evolution
particle swarm 'algorithm