Towards AI-based Distributed Lighting Control Systems

Abstract

The paper discusses an approach which is targeted at obtaining
maximal benefits of contemporary advanced lighting systems. The
benefits are expressed in terms of improved energy efficiency
(i.e. lower power consumption) or citizens quality of life. Applying
proposed solution one could use intelligent control methods which
functionality goes far beyond simple preset lighting scenarios as
109110
it is present in existing commercial systems. The main problem
tackled here is a high complexity of control algorithms related to
a size of a state space compound of lighting profiles, fixtures
working parameters and varying environment conditions. The pro-
posed method, designed for solving this issue, is using decom-
posable graph representations of the environment under control,
and multiagent system deployed on it. An important component of
the system is a rule-based engine, adapting lighting control parame-
ters to actual environment needs.