Vehicle interactive perception trajectory prediction system and method based on graph convolution

The invention discloses a vehicle interactive perception trajectory prediction system and method based on graph convolution. The prediction system comprises an input preprocessing module, a DGAC module and a trajectory prediction module which are connected in sequence. Collected vehicle data are input into the preprocessing module to generate a feature matrix and graph topological data, then multi-semantic graph spatial and temporal features are extracted through the multi-layer DGAC module, the attention mechanism is utilized to selectively pay attention to the difference between the spatial dimension, the time dimension and the channel dimension and key information, and finally the trajectory prediction module is utilized to predict the trajectory of the vehicle. And generating a predicted trajectory of the surrounding environment vehicle in a future period of time. According to the method, experimental verification is carried out through two real data sets, and experiments show that the model has higher gr