Insulator Pollution Grade Recognition Method and Platform

Abstract

Insulators play an important role in the power grids. However, when they get polluted, the flashover phenomenon may take place, which will bring huge losses to the power grid companies. The flashover phenomenon is closely related to the pollution levels of insulators, which need being measured to guide the maintenance work of insulators. If there is a convenient and accurate method to detect the pollution levels of insulators, the work to prevent the flashover phenomenon will be more efficient. To this end, an automatic insulator pollution grade recognition method based on hyperspectral imaging technology and SVM (Support Vector Machine) is proposed. What's more, this paper implements a matching automatic recognition platform with friendly human-computer interaction interface by PLC (Programmable Logic Controller) and Qt. The procedures of the automatic recognition method, the hardware composition and the software work flow of the automatic recognition platform are presented successively. At last, the final integrated recognition platform and the accuracy of the recognition method are exhibited to show their effectiveness.