Adaptive Decentralized Control Of Multi-machine Systems

Stabilization control of power systems is very important, and it is the key to maintain safe and reliable operation of power grids. One of the most economic and efficient methods of enhancing the operation stability is to improve the performance of power systems through modified control schemes.Therefore, the study of power system control design has practical demand and economic value.Multi-machine power systems are a class of large and complex dynamical systems with high nonlinearity, strong interconnections, changing operation conditions and uncertainties. These features make conventional control schemes hard to maintain the transient stability when the parameters and operation conditions vary sharply. The thesis is focused on adaptive decentralized control of large-scale power systems. The goal is to explore fully decentralized adaptive control schemes with nice transient process,guaranteeing the stability regardless of uncertain parameters and changing operation conditions. To this end, the following topics are studied:1. An adaptive decentralized tracking control scheme is proposed for a class of uncertain large-scale nonlinear systems generalized from multi-machine excitation control systems and steam valve control systems.The proposed control scheme is totally decentralized, and guarantees the global uniform boundedness of all signals, and makes the output track the desired trajectory. Numerical simulation results demonstrate the nice output tracking performance and robustness of the proposed controller.2. An adaptive decentralized excitation controller is designed to globally stabilize multi-machine power systems. Two smooth functions are introduced to counteract the effect of the unknown time-varying interactions, then a completely decentralized control structure is achieved. The proposed control scheme is tested on benchmark power systems. Various simulation results have showed better transient performances and enhanced the transient stability.3. An adaptive decentralized control scheme is proposed to globally stabilize the multi-machine steam valve control system. The proposed control method is performed on a three-machine power system, and simulation results show transient stability enhancement of the power system in the face of a large sudden fault and uncertain parameters.