Modelling of middle income residential lighting load profile using a universal estimator

The intricacies of the impact occupants have on lighting loads in residential buildings are not reflected in most practices use for light usage profile. This investigation involves the use of a universal estimator (Adaptive Neuro Fuzzy Inference System (ANFIS)) for middle income lighting load usage profile development, prediction and evaluation) for energy and demand side management initiatives. Natural light, Occupancy (active) and Income level are the three factors considered in this study. Trapezium membership function was applied during the training process of the ANFIS model due to historical light usage pattern (time of use) data. The result obtained after validation of the developed model using the investigative data showed a better correlation of fit and root mean square error in comparison with the regression model. The developed approach has the ability to give better lighting prediction accuracy in relation to non-linearity data and behavioural tendencies.