Positioning Performance Improvement of RTK Using Kalman Filter for Autonomous Vehicles

Abstract

Various researches for estimating the precise position of a vehicle are being actively conducted with increased interests in automotive vehicles and driver safety services. Currently, one of the most widely used technologies for estimating the position of a vehicle is a method using a Global Navigation Satellite System (GNSS). GNSS is one of the technologies to determine the outdoor position and uses the signals provided by artificial satellites to determine one’s position. However, due to different causes of error, general GNSS cannot estimate the precise position. To solve this problem, a Real-Time Kinematic (RTK) technology which can improve precision by using correction data provided by a base station has been developed. RTK can estimate a more precise position than GNSS. However, if the integer ambiguity cannot be precisely calculated, a precise position of the centimeter (cm) level cannot be estimated. In this paper, we propose a method for estimating a more precise position using the Kalman filter to improve the precision of the position estimated using RTK, and show the results of experiments.