【作者】 马永芳； 王兆耀； 凌志浩；

【Author】 Ma Yongfang~1 Wang Zhaoyao~1 Ling Zhihao~(1,2) (1 Automation Department of East China University of Science and Technology,Shanghai.200237;2 Key Laboratory of Advanced Control and Optimization for Chemical Processes,Ministry of Education,Shanghai.200237)

【机构】 华东理工大学自动化系； 化工过程先进控制与优化技术教育部重点实验室；

【摘要】 为了更好地对复杂故障进行诊断和排除,提出了基于最小二乘支持向量机(LSSVM)的故障诊断专家系统。LSSVM是SVM的一种改进算法,能够较好的解决小样本、非线性、高维数和局部极小等问题。将LSSVM嵌入专家系统中,实现知识自动获取和快速推理,专家系统则完成知识库的管理与维护和故障信息等工作,两者的优势结合提高了系统的综合诊断性能。最后,利用此系统对模拟的电梯故障进行仿真,结果准确一致。更多还原

【Abstract】 In order to get better diagnosis in complexity faults,a fault diagnosis expert system based least square support vector is introduced in this paper.LSSVM is a improved arithmetic of SVM,and advanced in solving the problems such as small sample,nonlinear system,high dimension number and local minimum points.Obtaining knowledge automatically and reasoning are achieved with LSSVM embedded in expert system.Management and maintenance of the knowledge base and correlative information is implemented by expert system.The system with the respective advantages of LSSVM and ES has improved diagnosis performance.The simulation shows the effective and validity of the fault diagnosis expert system based LSSVM.更多还原