Conditional colored petri net for modelling an automated gdm system

Abstract

When a house/building is able to perceive the environmental conditions and regulate or control a priori internal/external parameters accordingly, we said that it has an intelligent behavior. A Smart House/Building (SH) is a home-like environment which is able to perceive the environmental conditions, administrative state and users’ inputs in order to regulate, control or perform various tasks for achieving desired levels of comfort. Consequently a SH requires a distributed set of sensors, actuators and experts’ knowledge bases which should be specialized for specific tasks or problems domains. In this framework a Group Decision Making (GDM) system becomes paramount to choose the best decision among the different experts’ knowledge bases in order to supervise or control the security, administrative variables and users’ components. In this paper a Conditional Colored Petri Net (CCPN) is shown to be an useful technique to model and validate concurrency and distribution among the different experts and components of our whole system. A CCPN is jointly used with a group decision making (GDM) system to solve possible nondeterminism and ambiguity in the real time control environment of a SH when two or more experts might have different advice/conclusions for a particular combination of events. In this paper a set of experts’ knowledge bases are modelled and validated trough CCPN working with a GDM system. A physical example for SH supervision illustrates our approach.