Abstract: Targeted at the problems of short transmission distance and poor anti-interference ability of traditional communication technology,unreasonable and low water resource utilization of traditional irrigation,this paper constructs a precision irrigation system for orchards based on LPWAN Internet of Things and expert system. The system collects environmental information such as air temperature and humidity,soil temperature and humidity through the acquisition nodes,then uploads data packets to the cloud server through LoRa and NB-IoT networks.The expert system realizes water requirement prediction and irrigation decision-making and other functions based on the collected environmental data and the decision-making mathematical model,and feedback the decision-making results to the irrigation control module. In order to improve the prediction accuracy,the GA-BP water requirement prediction algorithm is introduced,and the mango orchard is used as the experimental object. The results show that the average packet loss rate of the system is 0.45% within 1 200 m,revealing that the data transmission is stable and reliable. The root mean square error and mean absolute error are 0.074 5 and 0.109 1 mm/d,respectively,which have higher prediction accuracy than the BP model. During the experiment,the relative humidity of the orchard soil is between 80% and90%,which meets the conditions required for the growth of mango trees. The system can realize real-time monitoring and precise irrigation of the orchard environment,and can provide a reference for further improving the effect of irrigation on fruit yield and quality.