In order to solve the problem of automatic generation control of traditional hybrid unified power system, a cascade PD-fuzzy PID controller was proposed.The scaling parameters of the cascade PD-fuzzy PID controller were optimized by using gray Wolf optimization and hybrid optimization based on teaching and learning optimization algorithm (GWO-TLBO).In the optimization process, the burst load interference of 0.01pu was added into region 1, and the absolute error of integral time was taken as the evaluation function.In the analysis of the dynamic performance of the system, the AC link and AC-DC link are considered.The robustness and sensitivity of the cascade PD-fuzzy PID controller are further verified in the case of random load variation and nominal parameter variation. The dynamic characteristics of the system are analyzed by using the time response index. By comparing the optimal parameters obtained by the proposed method with the existing results, it is concluded that the settling time of △f1, △f2and △Ptieusing the cascade PD-fuzzy PID controller is increased by 93.36%, 84.46% and 91.13% respectively compared with PID, and by 82.59%, 56.15% and 65.06% respectively compared with the fuzzy PID.The simulation results show that.The cascade PD-fuzzy PID controller is superior to other controllers, and solves the problem of automatic generation control of traditional hybrid power system, which can provide reference for automatic generation control of power system.