Most diesel generator speed control systems use traditional PID control. An iterative learning control (ILC) strategy that takes PID as the learning law is proposed. Making use of the characteristics of ILC of accurate mathematical modeling independent of the controlled system and fast tracking convergence, the diesel generator speed control system controller is designed. In MATLAB/Simulink, a simulation model of the diesel generator speed control system is established. Three basic working conditions are simulated and studied. The tracking characteristics of the ILC in the diesel generator speed control process are analyzed in detail, and the control effects of the proposed control and traditional PID control are compared. The results show that the controller based on ILC is not only suitable for the diesel generator speed control system, but also has better dynamic performance. A new way is provided for the control of diesel generators.