Traditional Dijkstra algorithm cannot adapt for the transportation network with traffic rule constraints during path planning. In order to solve this problem, based on the previous network models and algorithms, an improved Dijkstra algorithm with traffic rule constraints was proposed. It added new "to be selected state" and "renewable state" on the nodes, to solve the problem of nodes with traffic rule constraints. Meanwhile, grandfather node was introduced, thereby generating triples information of each node in the traffic network. Therefore, taking this as a backtrack basis, the shortest path from the starting node to the destination node could be obtained. This algorithm is not only applicable to the transportation network with traffic rule constraints but with low complexity. The correctness of the algorithm was verified through theoretical analysis. The effectiveness of the algorithm is verified through experimental tests carried out with actual traffic network in Chaoyang District of Changchun city and random added traffic rule constraints.