Multi-Agent Coordination of Thermostatically Controlled Loads by Smart Power Sockets for Electric Demand Side Management

Abstract

This paper presents a multi-agent control architecture and an online optimization method based on dynamic average consensus to coordinate the power consumption of a large population of Thermostatically Controlled Loads (TCLs). Our objective is to penalize peaks of power demand, smooth the load profile and enable Demand Side Management (DSM). The proposed architecture and methods exploit only local measurements of power consumption via Smart Power Sockets (SPSs) with no access to their internal temperature. No centralized aggregator of information is exploited and agents preserve their privacy by cooperating anonymously only through consensus-based distributed estimation, robust to node/link failure. The interactions among devices are designed to occur through an unstructured peer-to-peer (P2P) network over the internet.The architecture includes novel methods for parameter identification, state estimation and mixed logical modelling of TCLs and SPSs. It is designed from a multi-agent and plug-and-play perspective in which existing household appliances can interact with each other in an urban environment.Finally, a novel low cost testbed is proposed along with numerical tests and an experimental validation. Index TermsMulti-agent systems, demand side management, thermostatically controlled loads, online optimization, dynamic consensus, anonymous agents, distributed predictive control.The research leading to these results has received funding from the Italian Ministry of Research and Education (MIUR) with project "CoNetDomeSys", code RBSI14OF6H, under call SIR 2014.M. Franceschelli is with the

Multiagent Consensus Control Strategy considering Whole-Process Thermodynamic Characteristics of Air Conditioning Process

“…Method 2 (distributed consensus control strategy): the required proportion of load reduction is selected as the consensus variable, and the load reduction is in proportion to the original power consumption of the air conditioning load. is strategy follows the idea of the existing methods [13,14]. Method 3 (the proposed method) (distributed consensus control strategy): the incremental cost of air conditioning load is selected as the consensus variable, and the optimal allocation of the adjustment power can be realized.…”

Distributed Asynchronous Greedy Control of Large Networks of Thermostatically Controlled Loads for Electric Demand Response

“…This is justified by the fact most of the existing policies consider smart TCLs. Specifically, among [3]- [11], only [8] considers instead off-the-shelf TCLs, and thus there, the DSR is implemented by using networked smart sockets.…”

“…Specifically, TCLs monitor the microgrid frequency, then respond to deviations by decreasing (if the frequency is lower than a threshold) or increasing (if higher) their consumption by means of model-based considerations. Peak load shaving programs which aim to desynchronise the TCLs' consumption are [6]- [8]. Then, [9]- [12] study the so-called load-following programs, where the TCLs are coordinated such that the overall network demand tracks a given time-varying desired profile provided by the ESP.…”

“…TCL-DSR programs can be seen as large-scale constrained mixed-integer optimization problems [8]. Moreover, to keep computationally tractable the optimizations, it is common to approximate the dynamics of TCLs as LTI systems [8], [12], [14] thanks to which is predicted when the TCL will/would be switched on/off. Those predictions are however strongly influenced either by parameters deterioration or disturbances, such as water withdrawal, season changes, etc.…”

Compensator-based periodically intermittent control for nonlinear second-order multi-agent systems

“…In recent years, the research on multi-agent systems has received more and more attention, and it has a wide range of applications, such as unmanned aerial vehicles, 1 mobile robots, 2 thermostatically controlled loads 3 and sustainable agriculture. 4 Multi-agent systems consist of many systems called agents and the network structure used to connect the agents.…”