Title: Distributed Control for Smart Lights

Speaker: Prof. Kevin M. Passino, Ohio State University

Date/Time: Friday, September 10, 2010, 9:30 am Mountain Time - 8:30 am Pacific Time       

Location: CSRI/279 SNL/NM videoconferenced to 915/S145 SNL/CA

Brief Abstract: Smart lights seek to modulate their intensity based on whether light is entering the room, turning themselves down when there is significant ambient light, and turning up when clouds go by. They can also react to room occupancy changes. In this talk a simple  experimental testbed for the study of smart lights is introduced. This testbed has eight lights and eight sensors arranged in zones. It is shown that centralized algorithms can force the average and minimum light levels to track desired values. Moreover, it is shown that a genetic algorithm can be used to pick light inputs to optimize the attainment of lighting levels. Next, it is shown that a distributed proportional-integral controller can achieve efficient light operation even in the presence of a disturbance. A distributed voltage balancing controller is introduced and it is shown that it can achieve this same efficiency with fewer communications. Comparative analysis is performed for the methods and the effects of adding partitions between zones is studied.

CSRI POC: Dave Schoenwald, 505-284-6285 and Jean-Paul Watson, 505-845-8887