Robustness Assessment of Distributed OPF Under Communication Non-Idealities Using Cyber-Physical Co-Simulation Framework.

In recent years, distributed optimal power flow (OPF) algorithms have paved the way for the operation of distribution systems, where the distribution system is decomposed into multiple small areas each solving its own local optimization problem and sharing values of boundary variables with neighbors periodically. In contrast with the conventional centralized approaches, the performance of distributed algorithms is highly impacted by the non-ideal conditions of the communication infrastructures used for information sharing. Therefore, it is necessary to assess the robustness of distributed algorithms against real-time communication non-idealities. In this regard, this article investigates the performance of a fast distributed algorithm, named Equivalent Network Approximation method (ENApp), for three-phase radial distribution system under three different non-ideal communication scenarios: (1) erroneous data transfer, (2) additive noise effect and (3) communication delays. The non-idealities are modeled considering the physical aspects of two widely used wide-area network mediums, viz. power line communication and optical fiber. The performance characterization of ENApp is done using IEEE 123 bus test system in a HELICS-based cyber-physical co-simulation environment. This analysis provides an insight into the robustness of the ENApp algorithm in terms of the convergence time and success rate.