As there are constraints such as the time delay and generation rate constraint(GRC)in the automatic generation control(AGC)regulation process

the frequency regulation characteristics of power systems cannot be reflected by the AGC control strategy based on linear models. For the saturation and the time delay in the AGC time-delay system

a genetic optimization strategy of AGC cascade control system based on the proportional-integral(PI)stability domain is proposed. AGC system is equivalent to cascade control system made up of load frequency control loop and unit control loop. The Hopf bifurcation algebraic criterion and the time-delay system stability criteria are used to compute the stability domain of the inner and outer loops

respectively. It is proved that the parameters of the saturation and the time delay will affect the PI stability domains. Then the PI stability domains are converted into constraints of controller parameter optimization. Optimization comparative researches for the inner loop have shown that the integral of the absolute error(IAE)index has better performance in suppressing disturbances while the integral of the time squared error(ITSE)index has smaller fluctuations for the outer loop. Genetic optimization results show that the proposed control strategy is able to effectively resist the impact of the saturation and the time delay on the AGC system. This work is supported by National Natural Science Foundation of China(No. 61074104).