In the past, intelligent irrigation systems mostly used first-level fuzzy control and expert experience was difficult to combine various environmental factors, resulting in small application scope and inaccurate irrigation. An intelligent irrigation system based on improved Elman neural network and fuzzy control was designed. According to the type of crop and planting time, the optimal soil moisture is obtained through the first-level fuzzy controller; according to the humidity difference and evapotranspiration, the irrigation decision result is obtained through the second-level fuzzy controller, so as to achieve the purpose of precise irrigation. The traditional evapotranspiration calculation method is more complicated, and the Elman neural network has the characteristics of strong stability and good dynamics. The IPSO algorithm is used to optimize it to solve the problem of easy to fall into the local optimum and slow convergence. There are good results in the prediction of crop evapotranspiration. Experiments show that the system runs stably and can achieve precise irrigation.