Predicting the cooling energy requirement for an inland building in South Africa

Energy consumption contribution of heating and cooling of a building has been on the rise in recent times. In this paper, the cooling demand of an inland building in South Africa is predicted. Hourly historical data of weather and cooling energy was collected from Ninja for the year 2019 and subsequently analysed using Julia programming language. Initially, the pattern of features is analysed to establish a picture of their behaviour. These features are then used to train and test a regression model. The analysis shows that the data collected depicts accurately the situation in the inland of South Africa for cooling power requirements. Results of the regression analysis show that there is a need to segregate the data to enhance the accuracy of the model.