电力变压器绝缘在线监测数据聚合模型及其应用研究

发布时间：2018-05-15 01:06

本文选题：变压器 + 绝缘状态　；参考：《昆明理工大学》2014年硕士论文

【摘要】：变压器作为电网中能量转换、传输的核心,是电网安全的第一道防御系统中的关键枢纽设备。电力变压器健康水平和运行状态的判断主要是依靠定期检修的方式,但这种方式却忽略了电力变压器的实际状态,没有根据检测量信息本身所表现出来趋势或者变化规律进行深入发掘,亦没有从“信息驱动”的角度来考虑,忽略了变压器潜在的状态变化,也就很难实现变压器的状态估计。绝缘是电气设备结构中的重要组成部分,随着电力系统电压等级的提高,绝缘已成为电气设备中的薄弱环节。电力设备绝缘遭到损坏引起事故时,电力系统就不能安全稳定地运行,也给国民经济造成巨大损失。变压器中最广泛使用的绝缘材料是油浸纸绝缘,油浸纸绝缘系统在运行中受到电、热、机械等多种作用下会不断老化。针对电力变压器面临着日益严重的绝缘老化问题,发生事故的概率不断增加,同时,由于电力变压器的绝缘状态评估受到多种内部劣化和老化因素的约束,且影响程度又不尽相同。因此,通过深入研究基于物联网的信息聚合架构及各种信息聚合算法,根据变压器的绝缘老化机理,采用相对劣化度来衡量变压器绝缘状态劣化程度,实现变压器绝缘状态的划分。研究了核主成分分析和支持向量机的变压器绝缘状态评估模型,并提出基于典型相关分析的变压器绝缘状态评估信息聚合模型,根据相关性分析的结果对变压器绝缘进行决策,确定变压器绝缘运行状态,应用实例分析表明该模型的有效性和优越性。基于典型相关分析的变压器绝缘信息聚合模型,融合了全局特征和局部特征,获得了较佳的变压器绝缘参量提取效果,达到了智能处理海量数据的功能,实现对变压器进行有效的绝缘状态评估,可以用来指导变压器的运行维护和状态检修,延长变压器的使用寿命,使变压器更为经济、高效运行。
[Abstract]:Transformer, as the core of energy conversion and transmission in power grid, is the key hub equipment in the first defense system of power grid security. The health level and operation state of power transformers are judged mainly by the way of regular maintenance, but this way ignores the actual state of power transformers. It is difficult to realize the state estimation of the transformer if the trend or the change law of the measurement information itself is not deeply explored, nor is it considered from the angle of "information drive", and the potential state change of the transformer is ignored, so it is very difficult to realize the state estimation of the transformer. Insulation is an important part of electrical equipment structure. With the improvement of power system voltage grade, insulation has become a weak link in electrical equipment. When the insulation of power equipment is damaged, the power system can not operate safely and stably, and it also causes huge losses to the national economy. The most widely used insulating material in transformers is oil-impregnated paper insulation. The oil-impregnated paper insulation system will aging continuously under the action of electricity, heat, machinery and so on. In view of the increasingly serious insulation aging problem faced by power transformers, the probability of accidents is increasing. At the same time, due to the variety of internal deterioration and aging factors, the insulation state evaluation of power transformers is restricted. And the degree of influence is different. Therefore, through the in-depth study of information aggregation architecture based on the Internet of things and various information aggregation algorithms, according to the insulation aging mechanism of transformers, the relative deterioration degree is adopted to measure the degree of insulation deterioration of transformers. The insulation state of transformer is divided. Based on kernel principal component analysis (KPCA) and support vector machine (SVM), an evaluation model of transformer insulation state is proposed, and an information aggregation model based on canonical correlation analysis is proposed to make decisions on transformer insulation according to the results of correlation analysis. The operation state of transformer insulation is determined, and the application examples show that the model is effective and superior. The transformer insulation information aggregation model based on canonical correlation analysis combines the global and local features, obtains a better effect of transformer insulation parameter extraction, and achieves the function of intelligent processing of mass data. It can be used to guide the operation maintenance and condition maintenance of transformers, prolong the service life of transformers, and make the transformers more economical and efficient.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM41

【参考文献】