Energy forecast using linear regression methods

• Main Author: Hanafi, Noor Hasnizam
• Format: Thesis
• Language: English; English; English
• Published: 2004
• Subjects: TK Electrical engineering. Electronics Nuclear engineering; TK1001-1841 Production of electric energy or power. Powerplants. Central stations
• Online Access: http://eprints.uthm.edu.my/7915/1/24p%20NOOR%20HASNIZAM%20HANAFI.pdf; http://eprints.uthm.edu.my/7915/2/NOOR%20HASNIZAM%20HANAFI%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/7915/3/NOOR%20HASNIZAM%20HANAFI%20WATERMARK.pdf; http://eprints.uthm.edu.my/7915/

Energy analysis and forecasting have always been the essential part of an efficient energy system planning and operation. This thesis presents the mathematical methods based on regression analysis and energy profiling for energy modelling and forecasting. Two applications of energy were analyzed such as energy heating demands and electricity demands. The methods are applied for the Energieversorgung Offenbach AG (EVO) in Gravenbruch and Offenbach at Germany, using 1995 and 2003 dat^ The models of energy heating demands were developed based on simple linear and multiple regression methods. The Mean Absolute Percentage Errors (MAPE) are compared between two models. Two approaches to determine the typical energy profile are proposed. The two approaches use similar outdoor temperature for energy heating demand profiles and for electricity demands using similar electricity profile. The proposed approaches were able to determine the typical profiles of different type of day.