基于自适应神经模糊法的PEMFC温度建模  

Temperature Modeling of PEMFC Based on ANFIS

作　　者：孙涛[1,2] 闫思佳[1,2] 曹广益[1,2] 卫东[1,2] 

机构地区：[1]上海交通大学燃料电池研究所,上海200030 [2]上海师范大学应用化学系,上海200234

出　　处：《计算机测量与控制》2005年第7期662-664,700,共4页Computer Measurement &Control

基　　金：国家863基金资助项目(2003AA517020)

摘　　要：质子膜燃料电池(PEMFC)工作被认为是21世纪最有希望的绿色发电技术,其原理涉及热力学、电化学、流体力学、传质学等理论,形成一个非线性复杂系统,难以建立数学模型;因此,利用模糊逻辑系统和人工神经网络具有为非线性系统建模的较强的逼近能力以及自学习能力,采用自适应神经模糊算法,建立PEMFC温度特性模型;利用测试数据作为训练样本,在氢气压力给定的条件下,以空气(或氧气)压力和冷却水作温度为模型的输入量,电池的工作温度为输出量,建立了3种不同PEMFC温度特性模型;表明该方法具有简单、可行、精度高等优点。并为PEMFC控制系统的设计和电池性能的优化提供了基本依据。The proton exchange membrane generation technology is considered as the most hopeful “green” power technology. The operating principles of proton exchange membrane fuel cell （PEMFC） system involve thermodynamics, electrochemistry, hydrodynamics and mass transfer theory, which comprise a complex nonlinear system, for which it is difficult to establisha mathematical model. Consequently,this paper utilizes the ability of the fuzzy logic system and neural network which can approach unlimitly and learn by themselves, and adopts the adaptive neuro-fuzzy algorithm, then establish the temperature model; and makes use of the test data as the training specimen, on the condition of the given hydrogen pressure, the air （oxygen） pressure and cooling water as input, the operating temperature of PEMFC as output, three different temperature model are established. The result of simulation indicates that the method have the merits of simplicity feasibility and high accuracy. And it provides the basic guidance for the design of PEMFC control system and the optimization of cell performance.

关 键 词：质子膜燃料电池(PEMFC) 非线性系统建模 自适应神经模糊推理系统(ANFIS) 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]