Automation on Enhancing the Ground Temperature Sensing Using Machine Learning Approach

Abstract

Geothermal Energy is the inexhaustible and clean form of energy derived from the earth’s crust for different purposes like the generation of electricity. Water or steam is then used to bring energy to the earth’s surface. Very shallow geothermal potential (vsgp) is an alternative resource for providing clean energy to rural and urban areas. The ground temperature is predicated on various factors such as weather and terrain variables such as latitude, longitude ground aspect and slope, monthly precipitation, sunshine duration, snow depth, and air temperature. The challenges faced are due to seasonal weather variations and at the depth of 1-2 m, significant changes occur in the ground temperatures. After the collection and pre-processing of the data, it can be used to train and deploy the Machine learning model. By using various algorithms such as Hyper-parameter testing, Random Forest, Decision trees, and Support Vector Machine, etc, we can predict the ground temperature at the unknown locations. By using the help of various accuracy functions such as the root mean square method (RMSE) or normalized root mean square error method (NRMSE) we perform the estimations and calculate the efficiency of the model. The model can also be used to show the ground temperature maps and can also the variations of the ground temperature as compared to the previous years and can predict the future ground temperatures. if the data pre-processing and the prediction is done accurately it can be implemented on a larger scale and thus increase the geothermal energy productions.