Diagnostics of Magnetron Transmitter System using Thermal Camera and Neural Network

Abstract

The magnetron transmitter system is one of the most important components of weather radar. Monitoring the condition of the system is crucial to ensure the radar is operating normally. In this study, the temperature monitoring of the magnetron transmitter system was carried out. Thermal camera MLX90640 accompanied by NodeMCU ESP32 microcontroller provides a temperature matrix, which can represent the overall temperature of the radar system. The diagnosis system is built based on a neural network consisting of a hidden layer, and an output layer that classifies six system conditions. The results of this study showed that the thermal camera can measure the temperature of the magnetron transmitter system. This system can obtain 100% accuracy in identifying the conditions of the magnetron transmitter system, namely magnetron on, magnetron off, radar off, modulator power supply overheat, switch array unit overheat, and signal processing faulty. This innovation is expected to be early detection of an anomaly that occurs in the magnetron transmitter system so that it can minimize downtime to make repairs due to the unpreparedness of spare parts.