A Study of Multi Layer Perceptron (MLP)

A Comparative Study on Machine Learning Models for Ultrasound-based Defective Insulator Identification

Abstract

Interruptions in the supply of electricity cause numerous losses to consumers, whetherresidential or industrial and may result in fines being imposed on the regulatory agency's concessionaire.In Brazil, the electrical transmission and distribution systems cover a large territorial area, and becausethey are usually outdoors, they are exposed to environmental variations. In this context, periodicinspections are carried out on the electrical networks, and ultrasound equipment is widely used, due tonon-destructive analysis characteristics. Ultrasonic inspection allows the identification of defectiveinsulators based on the signal interpreted by an operator. This task fundamentally depends on theoperator's experience in this interpretation. In this way, it is intended to test machine learningapplications to interpret ultrasound signals obtained from electrical grid insulators, distribution, class 25kV. Currently, research in the area uses several models of artificial intelligence for various types ofevaluation. This paper discuss the Multi Layer Perceptron(MLP) and Comparison of single and multipleMulti Layer Perceptron(MLP)