Turbine supercharged set is used to supply high pressure combustion-supporting air. But the control for combustion-supporting air flux is a control with the characteristics of time-varying and nonlinearity and complex coupling. It is difficult to get the good performance through the traditional control mode. Evaluating of the characteristics of fuzzy control with quick dynamic response, good robustness and neural network control with good study capability, good generalization capability, a fuzzy CMAC parallel controller based on GA is presented in this paper. The mathematic model of the air flux is presented; Simulation results show the validity, good stability and control performance of the proposed strategy.