| 摘 要: 为分析燃料电池系统特性，采用BP神经网络结构辨识质子交换膜燃料电池系统模型，模型输入为系统 实际输入，模型输出为电堆输出电压和电堆工作温度。由于PEMFC系统是一个时变非线性系统，采用一种串-并联前 向神经网络辨识结构模型，将模型前几个时刻输出作为模型输入，使得静态网络结构具有动态特性。BP网络模型通过 PEMFC系统所得到的实验数据进辨识。训练完成后BP网络模型输出与实际系统输出基本一致，结果表明BP网络模型能 够有效反映质子交换膜燃料电池系统输出电压和电堆温度特性。 |
| 关键词: 质子交换膜燃料电池；BP神经网络；非线性系统建模；模型辨识 |
|
中图分类号: TP183
文献标识码: A
|
| 基金项目: 国家自然科学基金(No.51375314). |
|
| PEMFC System Modeling Based on BPNN Identification Model |
|
KE Chao,GAN Yi,WANG Jun,ZHU Rongjie,CHEN Wei1,2
|
1.(1.University of Shanghai for Science and Technology, Shanghai 200093, China;2.
2.Shanghai1 Institute of Space Power-Sources, Shanghai 200245, China)
|
| Abstract: In order to analyze the characteristics of the fuel cell system,BP neural network structure is used to identify the PEMFC system model.The model input is the actual input of the system,and the model output is the output voltage and working temperature of the stack.Since the PEMFC system is a time-varying nonlinear system,a series parallel forward neural network is used to identify the structural model.Meanwhile,the output of the first few moments of the model is taken as the following model input,so the static network structure has dynamic characteristics.BP network model conducts identification with the experimental data of PEMFC system.After training,the output of BP network model is basically the same as that of the actual system.The results show that BP network model can effectively reflect the output voltage and the stack temperature characteristics of PEMFC system |
| Keywords: PEMFC;BP neural network;nonlinear system modeling;model identification |
