电力变压器机械状态诊断方法研究

Research on Diagnosis Method of Power Transformer Mechanical State

针对传统电力变压器诊断方法存在诊断精度和稳定性较低的问题,提出一种新的电力变压器机械状态诊断方法。通过SVM分类算法,设计一个能够提高电力变压器诊断精度和速度的算法程序;根据DTBSVM设计多分类器,分清各样本之间的分离最优解,并依据向量投影线性分布和非线性分布两种思路计算向量投影中心点,设计一套基于DTBSVM的故障诊断算法流程。实验结果表明,基于DTBSVM的电力变压器机械状态诊断方法提高了诊断精度与算法的优化程度。

Aiming at the problem of low accuracy and stability of traditional diagnosis method for power transformer,a new diagnosis method for mechanical state of power transformer is proposed.Based on the DTBSVM,a multi-classifier is de-signed to separate the optimal solution of each sample,and a set of fault diagnosis algorithm flow based on DTBSVM is de-signed by calculating the vector projection center points according to linear distribution and nonlinear distribution.Experi-mental results show that the mechanical state diagnosis method based on DTBSV M achieves the expected goal,improves the diagnosis accuracy and the optimization degree of the algorithm.