Nonlinear delay system is commonly found in the industry, designing controller for this system is somewhat difficult. The authors pointed out some solutions by using neural RBF to identify and nonlinear compensation. Thereby, transforming nonlinear delay system to linear delay system with determined matrix A and B. In order to develop further result performed, this paper proposes a solution to design an optimal controller based on predictive control model for a nonlinear delay system after compensating nonlinear components. Through the selection of appropriate objective functions, it allows us to find out the optimal control signal, which is an explicit function. It can also find out the exact forecast range and level control of the system.