Intelligent vehicle trajectory prediction method based on environmental sensitivity

The invention discloses an intelligent trajectory prediction method based on environmental sensitivity. The method comprises the steps of searching trajectory data of the vehicle and preprocessing thetrajectory data; selecting an abscissa range and an ordinate range to establish a road network area; dividing the road network area into equal grids according to the granularity; setting a grid mapping function according to the granularity, and converting the trajectory data into a grid trajectory sequence; performing statistical construction to obtain a grid granularity matrix; constructing a neural network model based on environmental sensitivity; training a neural network model; predicting by using the model prediction track; and updating the neural network model. According to the method,the road network is divided into the fine-grained grids, the overall situation of the road network is abstracted by using the neural network, and the information is added into the trajectory prediction model, so that the traje