Calibration method of simulation data of virtual GPS and truth value in real world

Abstract

At present, the complexity of the research and development process of the intelligent connected vehicle and the diversity of research methods increase the difficulty of establishing, integrating and verifying of virtual simulation experiments. In view of the deviation of the virtual vehicle sensor in the simulation experiment to the restoration of the real environment, this paper proposes to use the multi-point positioning method for data calibration, use the least square method to estimate the unknown point position in complex situations, and according to the error size after calibration, specific evaluation criteria are given. Finally, the scene verification and simulation evaluation of the calibration method of the GPS simulation data in the virtual environment and the truth value in the real world were carried out. The results show that this method effectively improves the restoration degree of the virtual simulation sensor to the real environment.