【作者】 赵杰；

【导师】 荣雅君；

【作者基本信息】 燕山大学 ， 电力系统及其自动化， 2009， 硕士

【摘要】 电力变压器是电力系统中最重要的电气设备之一,也是导致电力系统事故最多的电气设备之一,其运行状态直接影响系统的安全性水平。及早发现变压器的潜伏性故障,保证变压器的安全运行,从而提高供电的可靠性,是电力部门关注的一个重要问题。因此,研究变压器故障诊断技术,提高变压器的运行维护水平,具有重要的现实意义。本文主要对电力变压器故障性质的诊断和故障部位的诊断方法进行研究。首先,针对变压器故障气体与故障类型的关系,建立结构为5-12-6型的变压器故障诊断人工神经网络模型。采用6种改进BP(Back-Propagation)算法及原最速梯度下降法对已建立的网络模型进行训练,使用训练好的网络模型对变压器样本进行故障性质的诊断。通过仿真比较了6种改进BP算法在变压器故障诊断中的性能优劣。其次,利用粗糙集理论对不完备的变压器故障信息决策表进行属性约简和规则的提取,得到决策表中隐藏的规则(即“if...then...”规则),利用此规则构建Petri网络模型,并利用Petri网进行故障诊断。将粗糙集理论与Petri网结合起来一同进行变压器故障部位的诊断。最后,通过故障实例验证了,基于人工神经网络对变压器故障性质的诊断和基于粗糙集理论和Petri网对变压器故障部位的诊断都能准确的、快速的得到诊断结果,达到预期效果。更多还原

【Abstract】 Power transformer is one of the most important electrical equipments in the electric system, and is also one of the electrical equipments resulting in most electric system accidents. Its operating state affects system’s security level directly. It is an important issue for electrical department to find the potential faults of the transformer, to keep it operating safely, and to improve the reliability or power supply. Therefore, it is of great realistic significance to study the fault diagnosis technology of transformer and to increase the operating and maintaining level of transformer.This paper applies the power transformer fault quality diagnosis and fault location diagnosis synchronously.Firstly, research of the relationship between fault types and dissolved gas, a type of 5-12-6 artificial neural network model for transformer fault diagnosis is established. Using of six improved back-propagation network algorithms and fast gradient drop algorithm to train the established artificial neural network, using of transformer data samples to the trained network model and processing fault quality diagnosis. Through the practical fault examples compare the fault diagnosis performance of six improved BP algorithms.Secondly, through rough set theory reduce the attribute of incomplete information table and extract rules, gain the potential and diagnostic rules (if…then…rules), and then based on these rules, the optimum petri nets are built, and through the petri nets calculate fault diagnosis. Rough set theory and petri nets are integrated for fault location diagnosis of transformer.Finally, the correctness and speediness of fault quality diagnosis of transformer based on artificial neural network and fault location diagnosis of transformer based on rough set theory and petri nets are validated by the practical fault examples. Meanwhile, fault diagnosis results can achieve the prospective effects.更多还原