基于超参数优化的SOFC系统故障分类方法  

SOFC System Fault Classification Method Based on Super Parameter Optimization

作　　者：樊洋 付晓薇[1,2,3] 李曦 FAN Yang;FU Xiaowei;LI Xi(College of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065;Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System,Wuhan 430065;State Key Laboratory of Materials Processing and Die&Mould Technology,Huazhong University of Science and Technology,Wuhan 430074;School of Artificial Intelligence and Automation,Huazhong University of Science and Technology,Wuhan 430074)

机构地区：[1]武汉科技大学计算机科学与技术学院,武汉430065 [2]智能信息处理与实时工业系统湖北省重点实验室,武汉430065 [3]华中科技大学材料成形与模具技术国家重点实验室,武汉430074 [4]华中科技大学人工智能与自动化学院,武汉430074

出　　处：《计算机与数字工程》2023年第4期775-780,948,共7页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:61873323);材料成形与模具技术国家重点实验室开放课题研究基金项目(编号:P2018-016);湖北省自然科学基金项目(编号:2017CFB506,2016CFA037);武汉市科技计划项目(编号:2018010401011292);武汉科技大学国防预研基金项目(编号:GF201912);智能信息处理与实时工业系统湖北省重点实验室开放课题项目(编号:znxx2018ZD01)资助。

摘　　要：固体氧化物燃料电池(Solid Oxide Fuel Cell,SOFC)是一种高效、绿色环保新能源,是解决能源与环境危机的重要途径与必然趋势。SOFC系统数据具有高维度、非线性结构及工作状态复杂等特点。为了提高SOFC系统使用寿命,论文提出了一种基于数据驱动的固体氧化物燃料电池系统多故障分类方法。首先,利用主成分分析进行故障信息提取;然后,对降维的特征空间,采用粒子群优化算法对支持向量机进行超参数寻优;最后,构建超参数优化支持向量机分类模型。实验结果表明,该方法能够准确、有效识别SOFC系统故障类型。Solid oxide fuel cell(SOFC)is a high-efficiency,green and environment friendly new energy source,which is an important way and inevitable trend for solving the crisis of energy and environment.The SOFC system data has the characteristics of high dimension,nonlinear and complex working conditions.In order to increase the service life of SOFC system,this paper proposes a multi-fault classification method for SOFC system based on data driven.Firstly,principal components analysis(PCA)is used to extract fault information.Then,to reduce the dimension of the feature space,particle swarm optimization(PSO)is used to optimize the hyper parameters of support vector machine(SVM).Finally,the classification model is constructed with the support vector machine based on optimal hyper parameters(PSVM).Experimental results show that this method can identify the failure types for SOFC system accurately and effectively.

关 键 词：故障分类 SOFC 主成分分析 粒子群优化算法 支持向量机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]