Governor Design for a Hydropower Plant with an Upstream Surge Tank by GA-Based Fuzzy Reduced-Order Sliding Mode

This paper investigates governor design by reduced-order sliding mode for a hydropower plant with an upstream surge tank. The governing system is made up of a tunnel, a surge tank, a penstock, a wicket gate and servomechanism, a governor, a hydro-turbine and a grid. Concerning the components of the governing system, their mathematic models are established. Then, these models are interconnected to simulate the governing system. From the viewpoint of state space in modern control theory, the governing system is partially observed, which challenges the governor design. By introducing an additional state variable, the control method of reduced-order sliding mode is proposed, where the governor design is based on a reduced-order governing system. Since the governor is applied to the original governing system, the system stability is analyzed by means of the small gain theorem. An genetic algorithm is employed to search a group of parameters of the predefined sliding surface, and a fuzzy inference system is utilized to decrease the chattering problem. Some numerical simulations are illustrated to verify the feasibility and robustness of the control method.