Assessment of indoor illuminance and study on best photosensors' position for design and commissioning of Daylight Linked Control systems. A new method based on artificial neural networks. (1st July 2018)

• Record Type: Journal Article
• Authors: Beccali, M.; Bonomolo, M.; Ciulla, G.; Lo Brano, V.
• Is Part Of: Energy. Volume 154(2018)
• Page Start: 466
• Page End: 476
• Publication Date: 2018-07-01
• Subjects: Indoor artificial lighting -- Energy efficient lighting -- Intelligent lighting control -- Artificial neural network -- Lighting measures reliability; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2018.04.106
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Artificial lighting systems have to ensure appropriate illuminance with high energy efficiency according to best design practice and technical standards. These aims can be tackled, by incorporating a Daylight linked control system. However, the system behaviour is strongly influenced by several factors and, in particular, by the sensors' position. Indeed, very often the illuminance on work-plane is not fully correlated with illuminance measured by the photo-sensor used to control the luminaires. This fact leads to wrong information for the Daylight linked control systems affecting its efficacy. The artificial intelligence of Neural Networks can be exploited to provide a method for finding good relationships between the illuminance on workplane and the one measured in another surface. Artificial Neural Networks are able to process complex data set and to give as output the illuminance in a point. By the use of measured values in an experimental set up, the output of several Artificial Neural Networks related to different sensors placements have been analysed. In this way it was possible to find the position of the photo-sensor associated to the best forecast of the workplane illuminance with a mean square error of 2.20 E −3 and R 2 of 0.9583. Highlights: Artificial Neural Networks can handle complex relationships describing lighting distribution in spaces. Artificial Neural Network methodology has been used with the purpose of finding the best sensor placement. Artificial Neural Network method has been used to provide good predictions of illuminance on workplane. Approach is useful to provide the most reliable input for Daylight Linked Control systems control function.