节点文献
直流偏磁下变压器振动机理与振动信号分析研究
Research on Transformer Vibration Mechanism and Signal Analysis under DC Bias
【作者】 段旭;
【导师】 杨永明;
【作者基本信息】 重庆大学 , 电气工程, 2013, 博士
【摘要】 变压器作为一种电网建设所必需的电气设备其运行的安全可靠性历来是人们关注的焦点。随着我国特高压电网的建设,电压等级的不断提高,变压器容量的不断加大,其内部的电磁力也日益增大,自1989年3月13日发生严重的地磁风暴在北美严重影响了电力系统的正常运行以来,地磁感应电流导致的直流偏磁问题一直是研究的热点。关于变压器的振动机理前人也开展了许多的相关研究,但是考虑直流偏磁下变压器振动机理的变化方面的研究开展甚少。因此本文就这一方面开展了研究。利用理论上数学公式的推导,电磁场-结构力场之间的耦合场建模,研制变压器振动监测系统,搭建实验平台开展相关实验,通过盲源分离技术对采集的变压器振动数据进行分离这几种研究手段和方法开展了具体的工作。论文的主要创新性工作包括以下几点:①考虑直流偏磁的影响,结合现有的变压器振动的相关理论,推导出了变压器铁心和绕组所受电磁力方程和位移方程。并以此为基础讨论了变压器铁心和绕组随直流偏磁量变化的规律。建立了考虑磁致伸缩影响的电磁场和结构力场的耦合场模型,通过此模型考察了发生直流偏磁时变压器内部磁通密度、机械应力分布、张力分布以及位移场分布的变化规律。②进一步考察电力谐波对变压器振动的影响,并为此建立了基于混沌电路的电弧炉模型,该模型能够很好的模拟电网中实际电弧炉内电弧电压和电弧电流的非线性特性,尤其是模拟了其中的混沌特性。以建立的仿真电弧炉模型作为谐波源,通过仿真考察了变压器铁心和绕组振动变化的规律,通过模拟真实环境得到的变压器振动数据更有说服力。③研制了变压器振动在线监测系统,这套在线监测系统采用了LabView作为开发平台,实现了通道信号实时显示、实时频谱显示、通道选择、触发选择、触发设置、数据存储等功能,特别是采用了新的数据存储方式,通过创建数据库来管理复杂监测任务。还实现了信号捕捉功能,使得瞬变信号不会消失,并且当变压器振动超过阈值时会自动记录波形。该在线监测系统不仅能完成本文的测量任务,还能作为一般的数字存储示波器使用。本文还搭建了实验平台,在此实验平台上完成本文所要要开展的各项工作,包括不同直流偏置量与变压器振动行为之间关系的研究,相同直流偏磁下不同负载和三相不平衡对变压器振动的影响研究等实验项目,填补了本课题方向上的研究空白。④提出了采用基于FastICA的盲源分离技术考察直流偏磁与变压器振动之间关系的的研究方法。并根据分离出的信号特征分辨出了铁心和绕组的振动信号,通过此方法分别考察了铁心和绕组振动随直流偏置电流变化的规律,相比目前本课题相关方向的研究都是以变压器器身振动信号作为研究对象来说,更能揭示变压器振动的本质。
【Abstract】 The operation safety and reliability of transformer which is a kind of the necessaryelectrical equipment in grid construction have always drawn people’s attention. With thedevelopment of ultra-high-voltage (UHV) power grid, the inside electromagnetic forceof transformer is growing with the unceasing enhancement of voltage level and theincreasing capacity of transformer. DC bias caused by geomagnectically induced currenthas been a research hot spot constantly since March13,1989when a severegeomagnetic storm in North American affected the normal operation of power systemseriously. Scholars have carried out a lot of related researches about the transformervibration mechanism. But, few research concerns the change of transformer vibrationmechanism under the DC bias affect. Thus, in this paper, researches have been carriedout in this aspect. Theoretic mathematical formula derivation, electromagnetic-structurecoupling field modeling, several research means and methods like developing thetransformer vibration monitoring system, building experiment platform to carry out therelated experiments, separating the transformer vibration data acquired fromexperiments through blind source separation technology were adopted in this paper.The main innovation work of this paper includes the following points:①Considering the DC bias affect and combined with the existing transformervibration related theory, the electromagnetic force equations and displacement equationsof transformer coil and core have been deduced in this paper. Based on this, change ruleof coil and core following the DC bias value are discussed. Considering the influence ofmagnetostriction electromagnetic-structure, coupling field modeling is built. Based onthe model, inside magnetic flux density of transformer, mechanical stress distribution,tension distribution and displacement distribution change rule under the DC biasinfluence are investigated.②In order to further study the influence of the electric power harmonic totransformer vibration, an AC arc furnace model based on chaos circuit is set up. Themodel can well simulate the nonlinear characteristics of practical electric arc voltageand arc current in grid. Especially, it simulates the characteristics of chaos. Using thesimulated electric arc furnace model as harmonic source, winding and core vibrationchange rule under harmonic affect is investigated through simulation. The vibration dataobtained by simulating real environment is more convincing. ③The transformer vibration on-line monitoring system is developed in this paper.LabView is selected as the development platform. The functions like Channel signalreal-time display, real-time spectrum display, channel selection, choice of trigger, triggersettings, and data storage, etc. are achieved in the system. Especially, a new data storagetechnology is applied. Complex monitoring tasks is managed through creating adatabase. The on-line monitoring system can not only accomplish the measurement taskin this paper, but can also be used as a general digital storage oscilloscope. Experimentplatform is built in this paper, on which the work carried out in this paper is completed,including the relation study between the different DC bias value and transformervibration behavior, the affect of different load and three-phase imbalance to transformervibration under the same DC bias, etc.. This work fills the blank of the direction of thistopic research.④The research method of blind source separation technology based on FastICAis proposed in this paper to investigate the relationship between DC bias andtransformer vibration. According to the characteristics of isolated signal, coil and corevibration signals are identified. Winding and core vibration change rule under DC biasis investigated respectively. Compared with researches in the related subjects using thetransformer body vibration signal as research object, it well reveals the essence oftransformer vibration.
【Key words】 DC bias; transformer vibration; electromagnetic-structure coupling field; on-line monitoring; signal separation;