engleski

Analysis of Real-Time Kinematic Samples Error for Sensor Fusion Applications

Recent developments in the fields of autonomous vehicles led to their widespread acceptance in many usage scenarios which include remote inspection, particularly in places not easily reachable or dangerous for humans. Global navigation satellite systems have become popular due to their convenience and low cost. The main disadvantages of typically used GNSS systems are relatively low refresh rate and low positional accuracy. The real-time kinematic method can be used to enhance the precision of GNSS but it still provides a relatively low refresh rate. Inertial navigation systems are based on inertial sensors, such as accelerometers and gyroscopes which provide high refresh rate. The main disadvantage of INS systems is the inherently present positional integration error of inertial sensors. The sensor fusion approach can be used to couple both the high positional accuracy brought by RTK and the high refresh rate resulting from the use of INS. However, coupling RTK and INS via sensor fusion approach is not a straightforward process due to challenges introduced by outliers and discontinuities in RTK positional data due to various factors affecting the GNSS signal reception. In this paper we present a static analysis of captured RTK data that can be used for further development of a method for compensation of unwanted side effects to enhance and stabilize the positional accuracy of the sensor fusion systems that incorporates both RTK and INS approaches.

GNSS ; RTK (real-time kinematic) ; sensor fusion ; inertial sensors ; precision positioning

nije evidentirano