Doctor of Philosophy (PhD)
Electrical Engineering & Computer Science
An output-feedback decentralized optimal controller is proposed for power systems with renewable energy penetration. Renewable energy source is modeled similar to the classical generator model and is equipped with the unified power flow controller (UPFC). The transient performance of power system is considered and stability of the dynamical states are investigated. An offline decentralized optimal controller is designed that utilizes only the local states. The network comprises conventional synchronous generators as well as renewable sources with inverter equipped with UPFC. Subsequently, the optimal decentralized controller is compared to the initial stabilizing controller used to obtain the optimal controller. An online decentralized optimal controller is designed for discrete-time system. Two neuro networks are utilized to estimate value function and optimal control strategy.
Furthermore, a novel observer-based decentralized optimal controller is developed on small scale discrete-time power system. The system is trained followed by least square rules and successive approximation. Simulation results on IEEE 14-, 30-, and 118-bus power system benchmarks shows satisfactory performance of the online decentralized controller. And also, simulation results demonstrate great performance of the observer and the optimal controller compare to the centralized optimal controller.
Wang, Boyu, "Decentralized Optimal Control With Application In Power System" (2018). LSU Doctoral Dissertations. 4665.
Available for download on Tuesday, July 02, 2019