基于改进RBF神经网络模型的SOFC性能预测方法

固体氧化物燃料电池(SOFC)测试存在费用高、实施困难以及耗时长等问题，因此，提出了一种基于径向基(radial basis function, RBF)神经网络的SOFC建模方法。首先采用数据驱动的方式利用RBF神经网络模型对电池中阳极、阴极、电解质厚度等微观结构对SOFC性能的影响进行分析，然后针对RBF神经网络模型参数选取困难、易陷入局部极值的问题，提出一种改进果蝇算法(improved fruit fly optimization algorithm, IFOA)对其进行优化，自动确定模型参数的同时确保其收敛于全局最优解。仿真结果表明，所提方法能够准确描述微观结构变化对SOFC性能的影响，相对于支撑向量机(support vector machine, SVM)模型能够获得更高的预测精度。

Performance Prediction Method of SOFC Based on Improved RBF Neural Network Model

There are some problems in the test and experiment of solid oxide fuel cell (SOFC), such as high cost, difficult implementation and long time-consuming. Firstly,radial basis function(RBF) neural network model is used to analyze the influence of microstructure such as anode, cathode and electrolyte thickness on SOFC performance by data-driven method. Then, an improved fruit fly optimization algorithm(IFOA) is proposed to solve the problem that the parameters of RBF neural network model are difficult to select and easy to fall into local extreme. IFOA is used to optimize the model, automatically determine the model parameters and ensure that it converges to the global optimal solution. The simulation results show that the proposed method can accurately describe the effect of microstructure changes on SOFC performance, and can obtain higher prediction accuracy than support vector machine (SVM) model.