Traffic state perception is an indispensable part in autonomous vehicles’ controlling. Meantime predicting the future trajectories of agents in the context is of great significance in perception of traffic states. However, challenges are that agents are not separated but interacting with other agents as well as the surrounding physical scene. Currently, most of the best models are based on graph neural network, but prior works normally include only position and timestep without other available and valuable information. Our model incorporates more information which conforms to the ever-developing V2I (Vehicle-to-Infrastructure) technology. More importantly, we have proposed a novel relative-position-wise attention mechanism and constructed the graph not only based on distance but on their yaw angle and velocity. In the main dataset used in the trajectory predictions, our model has achieved state-of-the-art performance. Also, we have contributed a simulated dataset named Shougang Dataset upon the map of Beijing Shougang Industrial Park. We have applied our model to Shougang Dataset and obtained an excellent performance in trajectory prediction.

暂无