基于粒子群算法的模糊大脑情感学习非线性系统辨识  

Nonlinear system identification based on fuzzy brain emotional learning with particle swarm algorithm

作　　者：孙园 曾惠权 欧阳苏建 高佳倩 王绮楠 林智勇 SUN Yuan;ZENG Huiquan;OUYANG Sujian;GAO Jiaqian;WANG Qinan;LIN Zhiyong(School of Electrical Engineering&Automation,Xiamen University of Technology,Xiamen 361024,Fujian,China;Xiamen Key Laboratory of Frontier Electric Power Equipment and Intelligent Control,Xiamen 361024,Fujian,China)

机构地区：[1]厦门理工学院电气工程与自动化学院,福建厦门361024 [2]厦门市高端电力装备及智能控制重点实验室,福建厦门361024

出　　处：《山东大学学报（工学版）》2024年第1期25-32,共8页Journal of Shandong University（Engineering Science）

基　　金：福建省自然科学基金资助项目(2020J01281);厦门市自然科学基金资助项目(3502Z20227215);厦门理工学院高层次人才科研启动项目(YKJ22060R);厦门理工学院研究生科技创新项目(YKJCX2021128)。

摘　　要：为提高神经网络模型在解决非线性系统辨识问题上的精度,提出一种基于粒子群算法的模糊大脑情感学习模型。该模型包含大脑情感学习网络,在利用系统历史数据对模型进行训练的基础上,通过适应度函数动态调整网络结构中的权重因子,提高网络学习效率和辨识精度。在对连续搅拌反应器系统辨识试验和对sin E强非线性对象逼近试验中,与常规模糊大脑情感学习模型、BP神经网络和RBF神经网络相比,本模型拥有更高的逼近能力和更快的收敛速度,解决了基于试错法导致的模型参数调整时间长、模型不稳定等问题,为辨识的实际应用提供了可行的模型选择。A fuzzy brain emotional learning model based on particle swarm algorithm was proposed to improve the accuracy of neural network model in solving nonlinear system recognition problems.The model contained a brain emotional learning network,and based on the training of the model using the system historical data,the adaptation function made the dynamic adjustment of the weight factors in the network structure to improve the network learning efficiency and recognition accuracy.In the identification test of continuous stirred tank reactor and the approximation test of sinE strong nonlinear object,compared with the conventional fuzzy brain emotional learning model,BP neural network and RBF neural network,this model had higher approximation ability and faster convergence speed,and solved the problems of long adjustment time of model parameters and model instability caused by trial-and-error based method,which provided a feasible model for practical application of recognition.

关 键 词：粒子群算法 类脑神经网络 大脑模糊情感学习模型 神经网络系统辨识 非线性系统 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]