Electric Power Load Forecasting

Problem Statement

Using time series analysis, can we forecast short-term power load demands and detect system disruptions based on historical load data in conjunction with weather data early enough to minimize economic impact?

Background

Why forecast load?

Load forecasting is the process of predicting the future of load demands in electrical systems using historical data. It is an essential component for planning in the electricity industry as it plays a vital role in power management and electric capacity scheduling. Electric power load forecasts can also serve as a key indicator in anomaly detection which could signal potential disaster in the area. This could be valuable information for electrical distribution companies.

In our project, we set out to engineer a model that will help us forecast short-term forecasting with reasonable accuracy.

Please refer to the pdf report for details.