基于BP神经网络的燃料电池发动机温度模型

Temperature Modeling of PEMFC Based on BP Neural Networks

燃料电池堆的热特性对燃料电池整体性能和寿命有重要影响,电堆温度特性具有不确定性和非线性,本文利用BP神经网络建立了电堆温度模型,并通过实测数据分析了神经网络模型的特性,研究结果表明,神经网络可以用于电堆温度模型的建立,为质子交换膜燃料电池电堆的建模与控制提供了一条可供参考的途径.

The thermal properties of a fuel cell stack have great influence on its integral behaviors and lifespan.The temperature characteristic of fuel cell stack is uncertain and nonlinear.The thesis built a model of the stack temperature with BP neural network and analyses the properties of the BP network with test data.The results showed that neural network can be used for modeling of the stack temperature.The approach is referencable for the modeling and control of PEMFC stack.