【作者】 章剑光；

【导师】 周浩；

【作者基本信息】 浙江大学 ， 电气工程， 2004， 硕士

【摘要】 本文以变电设备状态检修应用研究作为课题,围绕状态检修策略的三大组成部分:状态信息采集、状态诊断方法和检修策略展开应用研究工作: ① 阐述设备状态量获取方法,研究设备状态量信息的分类整理及设备性能指标的划分方法。 ② 神经网络算法在变电设备故障诊断模型的应用研究。采用面向对象的Delphi编程工具,建立基于带动量因子的标准BP、Bold Driver、SuperSAB和RPROP四种神经网络算法模型,并应用于主变压器溶解气体故障诊断分析(DGA)。通过算法建模及性能比较,选择更优的主变压器DGA故障诊断方法。 ③ 变电设备健康状态统一评价模型应用研究。结合专家系统、神经网络算法以及对不确定知识的模糊数学处理,最终得出设备健康状态的综合评价结果。 ④ 变电设备检修策略应用研究。建立可靠性维护(RCM)策略模型,综合RCM重要性指标及设备状态指标,运用于变电设备维护策略管理实践,建立更加科学、合理、经济的检修维护计划。 ⑤ 结合工程实际阐述本人所从事的绍兴电力局变电设备状态检修分析系统的开发应用情况。更多还原

【Abstract】 This thesis focuses on the condition based maintenance (CBM) on transformation equipment. With the three key components of the condition inspection analysis: acquisition of condition information, assessment of condition information, and application of overhaul tactics, the main works are as follows:① The expatiation of condition variables collection and the methods of classification on condition variables and the partition on performance index.② Research of the application of artificial neural network (ANN) on transformation equipment fault diagnosis model. Presented in this paper, we apply an evolutionary ANN programming based on versus standard BP, Bold-Driver and SuperSAB and RPROP algorithm. Our comparative analysis will find a better model on dissolved gas oil (DGA) fault diagnosis.③ Research of the uniform model on the condition information assessment. By using the expert system (ES), ANN and fuzzy mathematics, we give the general evaluate result of equipment.④ Research of the maintenance strategy on transformation equipment. With the Reliability Centered Maintenance (RCM) strategy, we balance between the condition indexes and importance indexes, and give a better maintenance plan.⑤ Introduce the CBM information system of Shaoxing electric power company which we have designed.更多还原