Controller Hardware-in-the-loop Implementation for Agent-based Distributed Optimal Power Flow Using ADMM on Cyber-Physical Microgrids

摘要

This paper presents the optimal power flow (OPF) implementation in the distributed manner of a cyber-physical connected microgrid. A laboratory testbed is introduced for the deployment including digital real-time simulator OPAL-RT, a cluster of Raspberry PIs (RPIs) and a local communication network. The general OPF problem is formulated and separated into bus-based subsystems. Agent is run on each RPI to implement alternating direction method of multiplier (ADMM) to solve OPF based on local information and adjacent data, then send power set points to physical system on OPAL-RT. The experiment on a 6-bus microgrid test case shows that the system operates at the optimal state under the control of the cyber system.