Development of pattern recognition based ANN for energy auditing and inefficiency diagnostics of influential design elements utilising electrical energy data

The objective of this study is to develop a pattern recognition based artificial neural network (ANN) for energy auditing and inefficiency diagnostics of the most influential design elements in buildings. The influential design elements examined include: envelope insulation, glazing insulation, glazing size, infiltration, economiser, building direction and orientation, and daylighting control. For generating the data bank needed for training the ANN, several buildings with different floor areas and known inefficiencies are simulated and their energy performance patterns, in terms of monthly energy demand and consumption, are determined. Monthly energy performance data including energy demand and consumption information for one year as well as building floor area are input. The developed ANN is validated and it is found that the expert auditing tool developed is effective for diagnosing the inefficient design elements. The application of the developed algorithm for real buildings and actual data is recommended for future work.