A kind of PID neural network (PIDNN) control strategy based on GA-BP algorithm is presented to improve the performance of the routine PID on nonlinear object. The PID control law is incorporated with the neural network to produce the PIDNN controller, whose parameters are optimized by the GA-BP algorithm. In order to verify the validity of the PIDNN controller and the corresponding PIDNN control strategy, the PIDNN controller is adopted to carry out simulation researches to a kind of nonlinear system. The simulation results indicate that the GA-BP algorithm convergence speed is fast, its control stability and speed of the PIDNN controller based on GA-BP algorithm are obviously better than the routine PID controller.