USF inventors have developed multi-agent decision making systems and methods based on primal-dual decomposition-based distributed optimization that are designed for economic operation and frequency regulation. The effect of these decision making strategies on power system frequency responses and coupling dynamics between the power market and system were optimized through rigorous testing in dynamic simulation platforms modeling real-world implementation issues. Test results show that the invented architectures are guaranteed to converge to the optimal power flow and pricing solution in less than 200 seconds for meshed networks of multiple areas and generators while regulating the frequency within adjustable bounds at all times. Furthermore, the architecture operates on limited non-sensitive information exchange among autonomous entities, which also reduces operational burden.
Competitive Advantages:
• Straightforward radial and scalable mesh network implementation • Discreet information exchange between community and utility • Convergence guaranteed stable design • Economic computing/communication load
(Left) Sub-Gradient Based and (Right) LUBS-Based Architectures
Desired Partnerships: