Power demand forecasting for district building

Abstract

The study of electricity demand forecasting has been of interest to researchers and organizations; in recent years, they also focus on building load forecasts. Researchers, power companies, software companies have developed many models. However, the results of accurate electricity demand forecasting remain challenging. One of the predictive methods is the use of the ANN network, which is used by many researchers with many positive results but continues to be studied and developed. Accurate electricity demand forecasting applied to among Demand Response, Demand Side Management, Energy Efficiency program, Micro Smart Grid, and Smart Building will help reduce the cost of electricity for the consumer. Accurate forecasting is also beneficial for electricity distribution and transmission companies and power systems operators. In the project district building Microgrid in Downtown Montreal, a content that needs attention is the forecast of the demand for electricity in the district building. In this thesis, we will review previous studies, synthesize substances related to the significant problem - solutions, research directions in electricity demand forecasts for buildings. Based on the implemented model present in this thesis, the short-term load forecasting model is proposed for a campus using the neural network model. The best suitable model was found when using many compare models and combined between them.

Date	3 Jun 2020
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Louis-A. Dessaint (Supervisor)