In order to provide real-time and accurate path query service and realize the path planning of moving objects, a spatio-temporal constrained path query method based on dynamic pruning P-region is proposed, which takes road network and moving objects as research objects and takes time and space as query conditions. The query pruning problem of spatial index is transformed into a fast query problem of non-Euclidean spatial distance, and the data preprocessing, index design, road network drawing and path query are analyzed in detail. The experimental results show that the dynamic pruning algorithm can significantly improve the query efficiency on the premise of guaranteeing query accuracy.