固体氧化物燃料电池的数学模型及自适应神经模糊辨识模型的研究

Modeling of Solid Oxide Fuel Cell Based on Mathematical Theory and Adaptive Neural Fuzzy Inference System Identification

固体氧化物燃料电池(solid oxide fuel cell,SOFC)是21世纪最有生命力的发电技术之一。文章从SOFC实际应用的角度出发,应用改进的自适应神经模糊推理系统(adaptive neural fuzzy inference system,ANFIS)对SOFC建立了负载稳定和负载变化2种情况下的电特性模型。由于数据来源不足,首先根据SOFC的工作原理,运用电化学、流体动力学等学科理论,建立SOFC的数学模型,基于该数学模型获取ANFIS辨识模型的训练和预测数据。仿真结果显示了改进的ANFIS技术对SOFC系统的建模和控制具有一定的实用价值。

The solid oxide fuel most promising power generation cell （SOFC） is one of the technologies in the 21st century. Starting from the practical application of SOFC and by use of improved adaptive neural fuzzy inference system （ANFIS）, the authors establish electric characteristic model for SOFC stack under the conditions of stable loads and variable loads. Due to insufficient source of the data, firstly according to the working principle of SOFC, by using the theory in electrochemistry and hydrodynamics, a mathematical model of SOFC is built; then on this basis the training and forecasting data for identification model of ANFIS is obtained, Simulation results show that the improved ANFIS technique is available for the modeling and control of SOFC.