An energy-efficient controller for shading devices self-adapting to the user wishes

In order to improve the user acceptance of an automatic shading-device controller, user wishes concerning the blind position are learned and integrated in the automatic controller through an innovative adaptation system developed with the use of genetic algorithms (GA). The adaptation process is composed of three distinct modules: a wish pre-processing module that prevents from learning energetically bad user wishes, the GA engine module and a sensitivity filter that ensures no parameters are needlessly changed. The system has been tested with several sets of synthetic wishes and shows a powerful ability to find satisfactory solutions, even with complex combinations of wishes.