Integrated design and real-time implementation of an adaptive, predictive light controller

Daylight–electric light integrated schemes encompassing soft computing models have been perceived as a lucrative option for lighting energy conservation. This paper exploits the quintessence of design and real-time implementation of an adaptive predictive control strategy for robust control of a daylight–electric light integrated scheme. To elicit daylight variations, occupancy detection and user preferences an online self-adaptive, predictive control algorithm is structured for real-time control of electric lights and window blinds. The adaptive, predictive model entails integration of an online, adaptive daylight illuminance predictor in conjunction with an electric light intensity control algorithm for interior illuminance regulation and a fuzzy-logic based window blind control algorithm to eliminate glare and solar heat gain. The control algorithm modelled with real-time sensor information administers an online process of identification, prediction and parameter adaptation. The prototype controller is successfully implemented in a test chamber. A real-time user-friendly simulator provides an online visualisation of illuminance performance indicators and control of the process. The anticipated synergetic effects of the online control algorithm validated in the test chamber highlights the benefits of the scheme in terms of glare control, illuminance uniformity and energy efficiency.