PEMFC的Elman神经网络建模与模糊神经网络控制

从质子交换膜燃料电池（PEMFC）实际应用的角度出发，采用Elman动态神经网络对PEMFC系统进行建模，以实验中采样到的PEMFC系统的工作温度输入输出数据训练网络，并采用动态反向传播学习算法根据误差不断调整网络参数直至达到要求精度。设计了一种适应模糊神经网络控制器，根据经验确定了初始隶属度函数和模糊规则，并采用自适应学习算法不断调整隶属度函数与模糊规则参数，使控制系统获得理想的输出。仿真实验以Elman神经网络模型为参考模型，使用自适应神经网络控制算法取得了较好的控制效果。总之，所设计的控制系统适合于控制PEMFC这样一类复杂非线性系统。

Elman neural network modeling and adaptive fuzzy control of PEMFC

To improve performance and durability of PEMFC, the operating temperature of PEMFC should be controlled within a specified range. In this paper, a Elman neural network identification model of PEMFC is developed based on the input-output sampled data. The process of identification is simple and cost less time. And then, an adaptive fuzzy neural network control algorithm is also designed. Finally using the Elman neural networks model of PEMFC, the simulation results of the control algorithm are presented. The results show the effectiveness of the proposed modeling and design procedures for PEMFC based on neural networks identification and the novel adaptive fuzzy control,