MODELING FOR EFFICIENT ELECTRIC ILLUMINATION

Abstract

Limited data is available in literature for obtaining relevant information in regard to performance of lamps. Hence there is a need to develop a model that can simulate performance characteristics of different types of lamps for varying conditions of applied voltage. In the present dissertation the objective is to investigate the suitability of using artificial intelligence based models for analysis of the performance of electric lamps. Two types of Artificial Intelligence methods are applied: (1) Artificial Neural Network model, (2) Fuzzy models. These models have been developed for three types of electric lamps namely Incandescent lamps, Fluorescent lamps, CFL lamps. The input parameters are the lamp ratings, applied voltage magnitude and lamp efficiency while the output is the consumed electric power & illumination flux. The models show performance comparable with experimental methods and can therefore be used as analytical substitute for experimental investigations.