A new approach on ai application for grounding resistor prediction in underground mines of vietnam

Abstract

Purpose. To apply artificial intelligence (AI) technology for predicting the earthing resistor of underground mines with consideration of climate change parameters. Methodology. In underground coal mines of Vietnam, the earthing system are everywhere equipped with individual rods combined with centralized grounding bed; this system significantly influences electrical safety and explosion safety. In daily operation, the resistor of the earthing system must be measured and tested regularly to ensure their value lower than the allowance limit (2). However, because of being affected by climate parameters (humidity and temperature in mines) this value varies frequently. By applying a new Neural Network with 3 hidden layers including variable parameters, the paper presents a new approach on predicting the earthing resistor. An algorithm is formed with visible and easily usable software for assisting the operator to predict the resistor. The prediction could be used for onsite management of a mine operator in the field of observing and testifying the earthing system in underground mines. Findings. Software is developed based on AI technology for assisting the operator to predict the value of the earthing resistor corresponding to climate change. Originality. Neural network with AI technology application is utilized relying on onsite measurements. Practical value. Prediction results could be used in case of difficulty in measurement. It will also help to correct or eliminate the measurement error from a mining technician.