Abstract: Based on the upper bound of time delay in networked control system(NCS) and the plant model,the whole system model of NCS is set up,and the generalized predictive pole placement control(GPP) is used in design of the controller, to ensure the system's validity and stability.Considering the randomicity of the network time delay,a compensation model for time delay error prediction is set up based on neural network,is used to compensate the output prediction value,then the predictive control algorithm for NCS based on the pole placement and the error compensation is constructed.The simulation results show that the method has higher precision and faster convergence speed.