Land-Vehicle Navigation System For Autonomous Driving With Averaged Nonholonomic Measurement

Abstract:

In this research, we propose a nonholonomic average speed measurement that newly recombines usable measurement in nonholonomic conditions to improve the heading accuracy of the vehicle Odometer/GPS/INS integrated navigation system. The average speed is less noise and the direction angle can be corrected more accurately because the information comes from data with long baselines. The advantages of this average speed is applied to nonholonomic measurements, which can improve the heading accuracy of the land vehicle navigation system. The proposed algorithm has been verified by Monte Carlo simulation. Verification results show improved heading accuracy and efficient applicability to navigation systems.