Abstract: The irrigation system has the characteristics of nonlinearity, multi-disturbance and time-lag, in order to achieve intelligent decision-making and accurate irrigation of the irrigation system, a Smith predictive variable universe fuzzy PID irrigation control model based on damping accumulated discrete grey prediction is proposed. Aiming at the shortcomings of fuzzy PID controller model, such as low control accuracy and weak adaptability, the exponential function expansion factor is designed to adaptively adjust the variable universe, and the Smith predictive compensator is used to eliminate the influence of time-lag, so as to improve the adaptability and robustness of irrigation control system. Combined with the prediction performance advantages of DGM and DAGM, DADGM is proposed to use damping trend parameters to slow down the change trend of process data, which effectively improves the stability and control accuracy of the irrigation system. Four control models of FPID, NVUFP, DGM-NVUFP and DADGM-SVUFP are constructed for irrigation simulation experiments. The results show that the steady-state error of DADGM-SVUFP is the best compared with other models, the settling time is 3.75 s and 1.29 s less than NVUFP and DGM-NVUFP, and the overshoot is 9.2% and 5.4% lower than NVUFP and DGM-NVUFP respectively. The irrigation test further verifies that the intelligent irrigation system based on DADGM-SVUFP has good adaptability, rapid response and high control accuracy. The control effect and system stability are better than other models, which can meet the intelligent decision-making and accurate control of the ox fertigation system.