Design of Embedded Intelligent Online Monitoring System of OLTC Based on Vibration Analysis

Abstract

Most of the traditional intelligent monitoring systems of on-load tap-changer (OLTC) are PC-based, which is expensive, high in power loss, complex in deployment, and large in space. A small, low-cost and low-power-loss embedded intelligent online monitoring system of OLTC based on vibration analysis is designed. The system uses an equal-integral-bandwidth feature extraction method of vibration signal based on fast Fourier transform (FFT), which solves the problems of difficulty in implementation and insufficient computing power in embedded systems. First, the FFT-based equal-integral-bandwidth feature extraction of vibration signal is discussed in detail. Then, a method applying fault diagnosis based on support vector machines (SVM) to the embedded system is proposed, and the framework of the intelligent online monitoring system of OLTC is given. Finally, the fault diagnosis experiments of OLTC are carried out. The experimental results show that after adopting the embedded system, it takes a very short time and the fault diagnosis is accurate, which can meet the needs of intelligent online monitoring of OLTC.