Data Analytics For Energy Management System: Energy Building Consumption

Abstract

Energy management system is a system that integrates people, technology, and science. Energy management system has strong relationship with energy consumption. The point of energy management system is not only methods to plan, organize, and control about energy consumption but also energy production for getting the goal. So, energy management system is needed in this industrial era for getting several objectives such as conservate, climate control, protection, and energy savings. This study will be focused on the process to establish energy management system in National Taiwan University of Science and Technology (NTUST). The Energy Management System (EMS) basically has very complex problem considering that EMS integrates people, technology and environtment. But, this study only will explain focused on energy consumption in NTUST and its forecast. Forecasting is very important part in energy management system because with forecasting we can predict the future usage of energy consumption so we can decide what should we do depend on the result. Forecasting is devided into several different method that can use for its suitable data pattern. This final report will explain how to forecast future events with various dataset with different method. The data in this study is retrieved from the database that record the energy consumption in NTUST. This study will explain about how to solve forecasting problem with smoothing method and Auto Regressive Integrated Method (ARIMA) with R software. Software is needed because there are many limitations if people should process the data manually. After processing the dataset with several step and different smoothing method, we can know that ARIMA is better smoothing method depend on the calculation of the error.