Based on analyzing the temperature dynamic model of Melting Carbonate Fuel Cell(MCFC),the non-linear of controlled object and the features of parameter distribution,the control scheme to sequential flow MCFC temperature control which based on genetic algorithm feedback multi-layer forward neural network is given out.The simulated results show that the control method is reasonable and effective.%通过对熔融碳酸盐燃料电池(MCFC)温度动态模型的分析,根据控制对象的非线性和参数分布的特点,提出了基于遗传算法的反馈型多层前向神经网络对顺流型MCFC温度进行控制的方案;通过对传统遗传算法的分析,提出了改进的遗传算法,用该算法对神经网络的权值进行优化,并给出了具体的学习算法。最后结合前人在MCFC模型动特性分析方面的工作和实验数据进行了仿真,结果表明这种控制方法是合理和有效的。
展开▼