基于改进差分进化算法的Wiener模型辨识

系统仿真学报 ›› 2016, Vol. 28 ›› Issue (1): 147-153.

基于改进差分进化算法的Wiener模型辨识

徐小平¹, 白博¹, 钱富才²

1.西安理工大学理学院,西安 710054;
2.西安理工大学自动化与信息工程学院,西安 710048

收稿日期:2014-07-29 修回日期:2015-01-27 发布日期:2020-07-02
作者简介:徐小平(1973-),男,博士,副教授,研究方向为进化算法,系统建模理论等。
基金资助:
国家自然科学基金(61273127); 陕西省教育厅专项科研计划项目(14JK1538); 陕西省自然科学基础研究计划项目(2014JM8325)

Identification of Wiener Model Based on Improved Differential Evolution (SADE) Algorithm

Xu Xiaoping¹, Bai Bo¹, Qian Fucai²

1. School of sciences, Xi'an University of Technology, Xi'an 710054;
2. School of Automation and information Engineering, Xi'an University of Technology, Xi'an 710048

Received:2014-07-29 Revised:2015-01-27 Published:2020-07-02

摘要/Abstract

摘要： 针对非线性Wiener模型的参数辨识问题,提出了一种基于Sigmoid函数及自适应算子改进差分进化(improved differential evolution algorithm with Sigmoid function and adaptive mutation operator,SADE)算法的参数辨识方法。利用Sigmoid函数及自适应变异算子改进了基本差分进化算法的变异操作部分,改进的方法能够有效地克服基本差分进化算法的过早收敛和不稳定性等缺点。将该改进差分进化算法应用于对非线性Wiener模型的参数辨识问题,达到了较高的辨识精度。在仿真试验中,与其它已有方法进行比较,仿真结果说明了所给的参数辨识方法是合理和有效的。

关键词: 差分进化算法, Sigmoid函数, 自适应算子, SADE算法, Wiener模型, 辨识

Abstract: To identify the parameters of nonlinear Wiener model, a new identification method was put forward based on a SADE algorithm. Its basic idea is as follows: the Sigmoid function and adaptive mutation operator were adopted to improve the mutation operation part of the basic differential evolution algorithm, accordingly, the disadvantages of the basic differential evolution algorithm, such as premature convergence, instability, etc, were effectively overcome. The proposed algorithm was used to parameter identification problem of Wiener model; moreover, the accuracy of identification was well improved. In the numerical simulation, compared with other relevant existing algorithms, simulation results show that the proposed method is rational and effective.

Key words: differential evolution algorithm, Sigmoid function, adaptive operator, SADE algorithm, Wiener model, identification

中图分类号:

TP391.9

徐小平, 白博, 钱富才. 基于改进差分进化算法的Wiener模型辨识[J]. 系统仿真学报, 2016, 28(1): 147-153.

Xu Xiaoping, Bai Bo, Qian Fucai. Identification of Wiener Model Based on Improved Differential Evolution (SADE) Algorithm[J]. Journal of System Simulation, 2016, 28(1): 147-153.

参考文献

[1] 萧德云. 系统辨识理论及其应用 [M]. 北京: 清华大学出版社, 2014: 56-68.
[2] 徐小平, 钱富才, 刘丁, 等. 基于PSO算法的系统辨识方法[J]. 系统仿真学报, 2008, 20(13): 3525-3528.
(Xu Xiaoping, Qian Fucai, Liu Ding, et al.Method of System Identification Based on PSO Algorithm[J]. Journal of System Simulation (S1004-731X), 2008, 20(13): 3525-3528.)
[3] 钱富才, 黄娇茹, 秦新强. 基于鲁棒优化的系统辨识算法研究 [J]. 自动化学报, 2014, 40(5): 988-993.
[4] 马海芳, 朱凌云, 王富强. 基于Wiener模型的pH中和过程非线性DMC控制[J]. 化工自动化及仪表, 2010, 37(9): 33-36.
[5] 刘长良, 任燕燕, 王东风, 等. 热工系统Hammerstein- Wiener模型辨识[J]. 计算机仿真, 2013, 30(9): 394-397.
[6] 胡德文, 王正志. 非线性系统Wiener模型辨识[J]. 自动化学报, 1991, 17(2): 151-159.
[7] 罗秋斌, 李秀英, 韩志刚, 等. 一种非线性系统多项式逼近的建模方法[J]. 系统仿真学报, 2008, 20(4): 829-832.
(Luo Qiubin, Li Xiuying, Han Zhigang, et al.Polynomial Approach for Nonlinear Modeling[J]. Journal of System Simulation, 2008, 20(4): 829-832.)
[8] 吴德会. 非线性动态系统的Wiener神经网络辨识法[J]. 控制理论与应用, 2009, 26(11): 1192-1196.
[9] Hatanaka T, Uosaki K, Koga M.Evolutionary Computation Approach to Wiener Model Identification [C]// Proceedings of IEEE Congress on Evolutionary. Piscataway, USA: IEEE Press, 2002: 914-919.
[10] 张燕, 李少远, 王笑波, 等. 基于粒子群优化的Wiener模型辨识与实例研究[J]. 控制理论与应用, 2006, 23(6): 992-994.
[11] Storn R, Price K.Differential Evaluation: a Simple and Efficient Adaptive Scheme for Global Optimization over Continuous Spaces[J]. Global Optimization (S1354- 1013), 1997, 11(4): 341-359.
[12] Wang L, Li L P.Fixed-Structure H^∞ Controller Synthesis Based on Differential Evolution with Level Comparison[J]. IEEE Transactions on Evolutionary Computation (S1089-778X), 2011, 15(1): 120-129.
[13] Onwubolu G, Davendra D.Scheduling Flow Shops Using Differential Evolution Algorithm[J]. European Journal of Operational Research (S0377-2217), 2006, 171(2): 674-692.
[14] 王凌, 钱斌. 混合差分进化与调度算法 [M]. 北京: 清华大学出版社, 2012: 112-121.
[15] 王永皎. 改进自适应差分进化算法求解大规模整数任务分配[J]. 计算机应用, 2012, 32(8): 2165-2167.
[16] Chang W D.Parameter Identification of Rossler’s Chaotic System by an Evolutionary Algorithm[J]. Science (S0036-8075), 2006, 29(5): 1047-1053.