Intelligent Multi-Agent Based Multilayered Control System for Opportunistic Load Scheduling in Smart Buildings

Today's power systems are subject to the high penetration of dynamic load. Volatility and intermittency of the dynamic load demand need to be compensated through optimization and scheduling without compromising user comfort. This paper proposes a multi-agent-based multi-layered hierarchical control system for residential load management under real-time pricing environment. The major objectives are to reduce peak load demand, electricity cost, and user discomfort. In doing so, different types of agents, i.e., price agent p(a), sensor agent s(a), decision agent d(a), load agent l(a), and action agent a(a), are developed to control residential loads, such as normal load (nl) and heavy load (hl). To handle price uncertainty, dynamically, optimal stopping rule (OSR) theory has been used. Two variants of OSR are proposed: 1) priority inversion logic-based OSR to subsidize the responsive consumers and 2) maximum energy consumption limit Q-based OSR-Q to maximize the profit of energy retailers. Finally, the proposed mechanism is validated on a set of loads to show the applicability and proficiency under a dynamic environment.