Analyzing Malicious Data Injection Attacks on Distributed Optimal Power Flow Algorithms

Abstract

This letter presents PowerModelsADA, an opensource framework for solving Optimal Power Flow (OPF) problems using Alternating Distributed Algorithms (ADA). PowerModelsADA provides a framework to test, verify, and benchmark both existing and new ADAs. This letter demonstrates use cases for PowerModelsADA and validates its implementation with multiple OPF formulations.

Assessing the impacts of nonideal communications on distributed optimal power flow algorithms

“…We therefore plan to analyze local termination criterion satisfaction for various distributed optimization algorithms and design alternate criteria that exhibit monotone behavior. Second, our contemporaneous work in [26] studies faults in the distributed optimization algorithm's computations. By combining ideas in [26] with the results of this paper, we aim to simultaneously address faults in both the distributed optimization algorithm's computations and the termination method.…”

“…Second, our contemporaneous work in [26] studies faults in the distributed optimization algorithm's computations. By combining ideas in [26] with the results of this paper, we aim to simultaneously address faults in both the distributed optimization algorithm's computations and the termination method. Third, we plan to extend the fault-tolerant method to consider asynchronous distributed optimization as well as dynamic communication networks where connections between agents may vary across iterations.…”

“…Although faults can occur in both the distributed optimization and the termination method, the proposed method focuses on faults in the termination method. We refer the reader to our other work [24]- [26] and the work in [27]- [31] on detecting data manipulation in distributed optimization algorithms. Assumption A5 implies that each agent shares the accurate status for its own local termination, but not necessarily the accurate termination statuses of other agents.…”