Learning-coordinate fuzzy logic control of dynamic quadrature boosters in multi-machine power systems

A learning-coordinated fuzzy logic control (LCFLC) strategy is investigated for the control of dynamic quadrature boosters (DQBs) installed distributively in a large-scale power system to enhance the power system stability. In this strategy a reinforcement learning technique is employed to search for optimal parameters of the fuzzy logic controllers according to a given performance index, to control the DQBs in a coordinated fashion. A simulation study is undertaken in a thirteen-machine power system. The simulation results show that the LCFLC can be used as an online control strategy to give coordination of the locally installed DQBs to improve the power system stability