Low-Cost Sensor Fusion Dead Reckoning using a Single-Frequency GNSS Receiver Combined with Gyroscope and Wheel Tick Measurements

A. Somieski, Ch. Hollenstein, E. Favey, C. Schmid

Abstract:	In recent years, many automotive navigation systems have been developed estimating the position and motion of a vehicle by combining GNSS measurements with data from other sensors, such as inertial sensors or odometers. The usage of measurements provided by attached sensors and/or the vehicle's CAN-bus not only improves the GNSS solution but also enables dead reckoning during periods of GNSS unavailability. This paper describes an embedded solution of Sensor Fusion Dead Reckoning (SFDR) utilizing a low-cost single-frequency GNSS receiver and MEMS gyroscope + wheel tick (GWT) information. On the basis of tick and gyroscope measurements from a single wheel the vehicle speed and heading rate can be calculated. In order to achieve optimal dead reckoning performance, the GNSS and sensor measurements are combined using a tightly-coupled extended Kalman filter allowing for continuous and automatic calibration of unknown sensor parameters, such as bias and scaling factor of the gyroscope and the scaling factor for the wheel ticks representing the wheel’s radius. Since the temperature dependent drift of the gyroscope bias during GNSS outages can degrade the dead reckoning performance significantly, the temperature and gyroscope bias are simultaneously measured during vehicle halt and a Temperature Compensation Table (TCT) is updated continuously. The performance of the presented low-cost “gyro + wheel tick” solution is evaluated by means of road tests carried out with a specially equipped test vehicle comprising a high-precision navigation system used as reference. Numerous road tests with different driving scenarios under varying environmental conditions were carried out, such as fast drive on a curved freeway, slow driving maneuvers in an underground parking garage, drive through tunnels and urban canyons. In addition, a large number of GNSS outages were simulated in order to study, improve and tune the dead reckoning performance as well as to analyze its accuracies statistically. The GWT results are also compared to the Differential Wheel Tick (DWT) Dead Reckoning technology, which derives the heading rate from the tick measurements of the left and right wheel. First statistical analyses show that the GWT dead reckoning solution performs more than twice as accurate as the DWT solution. In addition, GWT sensor fusion significantly outperforms a GPS-only solution, especially in difficult environments, such as urban canyons.
Published in:	Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010) September 21 - 24, 2010 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	1645 - 1652
Cite this article:	Somieski, A., Hollenstein, Ch., Favey, E., Schmid, C., "Low-Cost Sensor Fusion Dead Reckoning using a Single-Frequency GNSS Receiver Combined with Gyroscope and Wheel Tick Measurements," Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010), Portland, OR, September 2010, pp. 1645-1652.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In