燃料电池的模糊神经网络辨识建模与电压控制

针对直接甲醇燃料电池(DMFC)输出电压易受电池负载变化影响的问题,提出了利用模糊神经网络辨识技术建立DMFC的电特性模型.基于该辨识模型,设计了一个自适应模糊神经电压控制器,其参数采用改进的BP算法进行在线修正.仿真结果表明,对DMFC采用辨识建模的方法是有效的,建立的模型精度较高,所设计的自适应模糊神经电压控制器性能优越.

Fuel Cell Modeling Based on Fuzzy Neural Networks Identification and Its Voltage Control

An electric characteristic model of direct methanol fuel cell (DMFC) was established by using fuzzy neural networks identification technique to cope with the problem that the output voltage of DMFC is significantly influenced by the fuel cell load variation. Based on the identification model, an adaptive fuzzy neural networks voltage controller was designed. The parameters of the controller are regulated by adopting the novel BP algorithm. The simulation results show that the fuzzy neural networks identification modeling method is correct, effective and the model has better accuracy. Moreover, the performance of the designed adaptive fuzzy neural networks voltage controller is very superior.