基于改进ANFIS网络的绝缘子运行状态诊断研究  

Research on Diagnosis of Insulator Operating State Based on Improved ANFIS Networks

作　　者：张子蓬[1] 王淑青[1] 刘辉[1] 袁晓辉[2] 

机构地区：[1]湖北工业大学计算机学院,湖北武汉430068 [2]华中科技大学水电与数字化学院,湖北武汉430074

出　　处：《水电能源科学》2009年第4期194-196,227,共4页Water Resources and Power

基　　金：国家自然科学基金资助项目(50779020);湖北省自然科学基金资助项目(2008CDB395)

摘　　要：针对输电线路绝缘子性能的优劣对整个输电系统的安全运行起非常关键的作用,采用改进的自适应神经网络模糊推理系统(ANFIS)对绝缘子运行状态进行推理诊断,以有效诊断推理绝缘子污秽程度及闪络发生的可能性,训练中采用改进的学习方法,提高了推理网络的学习能力。试验结果表明,设计的绝缘子运行状态诊断系统具有优化学习能力,为准确推断绝缘子运行状态提供依据。Power transmission line insulator is an important part for power system security. Because insulator has complex operating environment and its infection factors interact on each other, the diagrosis of insulator running state is very difficult. It is needed to use some useful information to diagnose insulator operating state. ANFIS networks have strong knowledge expressing, fuzzy reasoning and learning ability. They are used to diagnose insulator operating state in this paper. In order to improve the knowledge gain ability for ANFIS, improved learning way is used to train reasoning parameter. Experiment results show that the designed ANFIS network has strong reasoning and learning ability, which can diagnose insulator operating state efficiently.

关 键 词：绝缘子运行状态 诊断 自适应神经网络模糊推理系统 改进的学习方法 

分 类 号：TM216[一般工业技术—材料科学与工程] TM852[电气工程—电工理论与新技术]