固体氧化物燃料电池的神经模糊控制策略研究  

Study on Neural Fuzzy Control Strategy of Solid Oxide Fuel Cell

作　　者：王倩如 王彩霞[1] 顾吉鹏 WANG Qian-ru;WANG Cai-xia;GU Ji-peng(College of Electrical Engineering,Northwest Minzu University,Lanzhou,China,730000;College of Information Engineering,Zhejiang University of Technology,Hangzhou,China,310000)

机构地区：[1]西北民族大学电气工程学院,甘肃兰州730000 [2]浙江工业大学信息工程学院,浙江杭州310000

出　　处：《热能动力工程》2022年第10期198-206,共9页Journal of Engineering for Thermal Energy and Power

基　　金：西北民族大学研究生科研创新项目(Yxm2021090);西北民族大学中央高校基本科研业务费项目(No.31920190067)。

摘　　要：固体氧化物燃料电池(Solid Oxide Fuel Cell,SOFC)具有多输入多输出、强耦合的特点,为了使其输出电压稳定设计了高效控制器,采用神经模糊控制方法对其输出电压进行控制。通过机理分析和实验数据拟合方法分别建立SOFC的机理模型和神经网络模型,在此基础上采用模糊控制策略对SOFC的输出电压进行控制,并应用神经模糊控制方法进一步提高了控制精度。通过MATLAB/Simulink仿真实验发现,SOFC神经网络模型得到的预测电压与实际电压之间的误差小于0.008 V,较其机理模型更加准确,所提出的控制策略能有效控制SOFC的输出电压。Solid oxide fuel cell(SOFC)has the characteristics of multi-input,multi-output and strong coupling.In order to design an efficient controller to stabilize its output voltage,this paper uses neural fuzzy control method to control its output voltage.The SOFC mechanism model and neural network model are established through mechanism analysis and experimental data fitting method respectively.On this basis,fuzzy control strategy is adopted to control the output voltage of SOFC,and neural fuzzy control method is applied to further improve the control accuracy.Through MATLAB/Simulink simulation experiments,it is found that the error between the actual voltage and the predicted voltage of the SOFC neural network model is less than 0.008 V,which is more accurate than its mechanism model.The proposed control strategy can effectively control the output voltage of the SOFC.

关 键 词：固体氧化物燃料电池 SOFC神经网络模型 模糊控制 神经模糊控制 

分 类 号：TK91[动力工程及工程热物理]