GPS/INS Integration Aided with Gyroscope-Free IMU for Pedestrian Applications.

摘要

This thesis investigates the use of a distributed configuration of accelerometers known as a gyroscope-free inertial measurement unit (GFIMU) as an aid to GPS/INS integration. While a GFIMU is theoretically capable of replacing a traditional strapdown IMU, there are several practical issues that make the approach less than ideal. Therefore, a combination of a GFIMU and a tri-axial gyroscope, referred to as GFIMU+ herein, is designed and tested. Inertial navigation with a GFIMU+ is called Gyroscope free inertial navigation + (GFIN+).;Methods used to de-noise inertial sensor measurements are discussed and implemented. The de-noised measurements are used to estimate the de-noised GFIMU+ estimates.;The performance of the GFIN+ system is evaluated in the field in different operating environments and different placements of the GFIMU. The results show that de-noising improves the results in most of the cases and an average improvement of about 30% horizontal RMS is seen as compared to GPS integrated with single MEMS based INS.;The primary piece of equipment required for the formation of a GFIMU+ is the associated configuration of accelerometers. Five MEMS IMUs were rigidly attached to a compact, DelrinRTM block for a prototype GFIMU+. The five tri-axial accelerometer measurements and tri-axial gyroscope measurements are combined in an Extended Kalman Filter (EKF) to obtain an estimate of angular velocity. The specific force and angular velocity estimates are inputs to the GPS/INS tight integration software.