MAS-Based Modeling of Active Distribution Network: The Simulation of Emerging Behaviors

Authored by Merkebu Z Degefa, Antti Alahaivala, Olli Kilkki, Muhammad Humayun, Ilkka Seilonen, Valeriy Vyatkin, Matti Lehtonen

Date Published: 2016

DOI: 10.1109/tsg.2015.2510547

Sponsors: No sponsors listed

Platforms: No platforms listed

Model Documentation: Other Narrative Flow charts Mathematical description

Model Code URLs: Model code not found

Abstract

Agent-based modeling of active distribution network helps to understand the dynamics and to design the control strategies for overall system efficiency. There is, however, a lack of generic and multipurpose agent definitions in existing studies. In this paper, a multi-agent system-based modeling of an active distribution network is presented using cooperative agents. A method to solve a network-wise objective of state estimation is explained with the proposed model. The network component agents are defined to be cooperative to meet the overall objectives and greedy to fulfil individual objectives such as energy cost minimization. A token-ring protocol is deployed for the agent communication among themselves, as well as with market and network operator agents. Furthermore, a MATLAB/Simulink model of active distribution network is used to simulate the emerging stochastic loading scenario, while the autonomous pro-sumer agents optimize their total energy cost responding to market price variations.
Tags
Agent systems State estimation Automation