Vehicle and pedestrian map track prediction method and device

The invention discloses a vehicle and pedestrian map trajectory prediction method and device based on a random walk graph neural network, and the method comprises the steps: determining a target city, and building a pedestrian and vehicle prediction training database based on the target city; constructing a random walk graph neural network model, extracting prediction training data from the prediction pedestrian and vehicle training database, and inputting the prediction training data into the random walk graph neural network model for training to obtain a convergent random walk graph neural network model; obtaining real-time target object data in the target city, and preprocessing the real-time target object data to form to-be-predicted target object data; and inputting to-be-predicted target object data into the convergent random walk graph neural network model, and outputting the occurrence probability of the to-be-predicted target object at the corresponding node. According to the embodiment of the invent