To achieve optimal control under all operating conditions for hydropower units, an intelligent robust controller (NSFLIRC) is proposed based on a novel state space equation and state-dynamic-measurement feedback linearization. Firstly, a nonlinear hydro-turbine regulation system (HTRS) model is constructed and a novel high-order state space equation model of HTRS is derived considering system tracking deviation. Then, the robust control for HTRS is achieved by combining the novel high-order state space equation and H∞ control. Furthermore, the slime mold algorithm (SMA) is improved by incorporating hybrid chaotic mapping functions, Levy flight, and a nonlinear weight coefficient. The intelligent optimization of H∞ controller parameters is realized using the improved SMA (ISMA). Finally, a nonlinear HTRS simulation platform is constructed to verify the performance of the proposed NSFLIRC. The simulation results show that when the overshoot is within a reasonable range, using NSFLIRC can respectively reduce the average values of adjusting time, rising time, overshoot, and power inversion by 46.6%, 29.06%, and 52.56% comparing to the traditional PID controller under six typical operating conditions; also, the NSFLIRC has low sensitivity to changes in system parameters and most operating conditions, and has strong robustness under conventional load changes and extreme three-phase short circuit conditions.