Distributed cyber-physical algorithms for wide-area control of power systems

With the number of Phasor Measurement Units (PMUs) in the North American power grid scaling up into the thousands, system operators are gradually inclining towards distributed cyber-physical architectures for executing wide-area monitoring and control operations using Synchrophasors. Traditional centralized approaches, in fact, are anticipated to become untenable soon due to various factors such as data volume, security against single point of failure, communication overhead, and failure to adhere to real-time deadlines. In this talk I will propose a distributed communication and computational architecture, and its associated distributed optimal control algorithms, for one of the most critical applications run by utility companies - namely, wide-area control of power flow oscillations following small-signal and large-signal disturbances in the grid. In this architecture, Synchrophasor data from PMUs located at the terminal buses of generators are first transmitted to the local control center for that particular balancing region, and thereafter to a set virtual computers in a local cloud computing network. Depending on the correlation between the generator states in terms of the modal content of the measured signals, a distributed communication strategy is chosen and established between these virtual computers in the cloud. A real-time model-predictive control design is then implemented using this distributed communication network, assuming synchronous communication. I will present simulations that highlight some interesting convergence and accuracy trade-offs for the real-time implementation of the wide-area controller, and also illustrate their architectural resiliency against denial-of-service and data-manipulation attacks using case studies from the recently federated DETER-WAMS testbed between NC State and University of Southern California.