Distributed Co-Simulation for Collaborative Analysis of Power System Dynamic Behavior

Abstract

Given the complexity, scale and heterogeneity of modern power systems, many comprehensive simulation tasks are almost impossible to carry out without collaboration from multiple institutions. One way to approach collaborative simulation is through co-simulation. In this approach each institution contributes a model and the simulation is run distributedly. Co-simulation environments can be centrally orchestrated or decentrally orchestrated, each having its own set of advantages and challenges. In this paper we analyse the merits and challenges of co-simulation for collaborative simulation between institutions through two simple co-simulation environments, one centralized and the other decentralized. We argue that a co-simulation environment for collaboration between institutions should provide functionality like standard co-simulation interfaces for a variety of simulation tools, remote management and con?guration, model compatibility checks, and failure detection and recovery. An environment with these characteristics can be easily adopted by a wide range of institutions, which would greatly aid in tackling the complexity of modern electrical power systems.