The paper aims to effectively integrating city parking resources and increasing the utilization of existing parking spaces. An intelligent parking prediction system based on NB-IoT technology is constructed. The system uses NB-IoT technology for information collection and transmission to realize the sharing of parking space information; The real-time changes of the parking space status information is considered, and the historical parking space occupancy data is used to establish the parking space prediction model, so as to infer the future short-term parking space change trend. Genetic algarithm(GA)-back propagation(BP) neural network parking space prediction model is established to improve the accuracy of parking space prediction. We take the historical data of an underground parking lot as an example, and the simulation results show that the predicted value of the prediction model is similar to the actual value and their trends are consistent， which can effectively and accurately predict the change of parking space state and has high precision.