Aiming at the control speed and accuracy of the AC servo system of the ship-borne rocket launcher, a composite control strategy combining RBF neural network based PID control with fuzzy control is proposed. The composite control strategy adjusts the weights of the two control methods on the controlled object according to the error information, so as to switch the two control methods without disturbance, and realize the continuous and stable output of the final control quantity. The simulation results show that the control strategy combines the advantage of RBF neural network based PID control with that of fuzzy control, and has high control precision, fast system response speed and small oscillation, which can better meet the performance requirements of the AC servo system of the ship-borne rocket launcher.