Prediction of Vehicle Trajectory Based on Fuzzy Colored Petri Net

Aiming at the difficulty in establishing model to describe vehicle running state, real road test was carried out to capture data which represent the movement of vehicles. Coordinate transformation method and transform relationship among different variable parameters were used to establishing represent model of vehicle movement. By using Petri net which had well layering and time sequence, vehicle trajectory, speed, side slip angle, and yawrate were treated as parameters to describe the movements of vehicle. Domain of discourse and subordinating degree function were confirmed, and fuzzy rules related to controllability and driving comfort were established. Verification tests results show that the Petri net model can describe the vehicle movement accurately, and the predict results of represent parameters were similarly with the real measured data.