Failure Prediction for High Voltage Induction Motor using Artificial Neural Network (ANN)

This study assesses the machine learning model used to analyze the most influential factor (MIF) of maintenance in rotating motor or induction motor and correlate the factors. There are many data collected in the industry but was not fully used to construct a decent application to detect the failure in the motor. A predictive maintenance approach can be applied to handle data-driven difficulties such as processing complexity, analysis time, and difficulty employing big data features for the maintenance of the motor. The objective of this study is to create an efficient machine learning model algorithm for HVIM failure probability. The study created the model using an artificial neural network (ANN) using three MIFs, vibration, temperature, and power factor, and correlate the model between the MIFs to strengthen the factor selected for maintenance. The model validated the MIFs, temperature, vibration, and power factor with their range of the motor operating between 38 °C to 41 °C, up to 3.8 of vibration, and 0 to 100 kW respectively. The failure prediction of the motor happens after the range of operating. Due to the analysis’s fitness, the HVIM maintenance prediction model based on MIF using ANN is suitable for industrial applications. As a result, it applies to industrial maintenance for future predictions and has the potential to be further improved in the future.