MEMS Inertial on an RTK GPS Receiver: Integration Options and Test Results

Inertial technology has evolved over the last fifty years from one based on iron gyros, with error control from zero velocity updates and star trackers to ring laser and fiberoptic gyros updated with GPS observations of various types. Today, micro electromechanical system (MEMS) inertial measurement units (IMU) are maturing to the state that low-cost commercialization is becoming feasible. At the same time, computer processor power is sufficient to allow both GPS and inertial data to be processed in a single processor. NovAtel Inc. and Honeywell have collaborated on the integration of the Honeywell HG1900 MEMS IMU and the NovAtel Inc. OEM4 GPS receiver. MEMS inertial and GPS data are collected and processed on the OEM4 resulting in a combined position, velocity and attitude solution that benefits from the synergies of the two technologies. The MEMS inertial navigation system (INS) provides continuous position when GPS satellite signals are not available, and reduces the time to obtain a real-time kinematic (RTK) position once GPS signals return. The objective of the integration is to provide continuous sub-meter level positioning (2 sigma) to address future requirements of the automotive industry. The system integration is described, particularly, the major components, the timing synchronization and the filtering methodology. In order to deal with some of the error characteristics of the MEMS system, various filtering methods were investigated before the requirements were met. In addition, various GPS observables were integrated in the filter, and the impact of various types of observables was tested. Test results are presented. In particular, a filter modification incorporating current and previous states was implemented, and the effect of adding GPS delta phase measurements to this implementation was quantified. The advantage of adding delta phase measurements is that the filter can take advantage of the low noise and small error growth of the phase measurement without the additional burden of maintaining ambiguity states. Finally, simulations were done to evaluate possible system performance improvements using distance measurement indicator (DMI) aiding.