An Adaptive Low-Cost GNSS/MEMS-IMU Tightly-Coupled Integration System with Aiding Measurement in a GNSS Signal-Challenged Environment

Abstract

The main aim of this paper is to develop a low-cost GNSS/MEMS-IMU tightly-coupled integration system with aiding information that can provide reliable position solutions when the GNSS signal is challenged such that less than four satellites are visible in a harsh environment. To achieve this goal, we introduce an adaptive tightly-coupled integration system with height and heading aiding (ATCA). This approach adopts a novel redundant measurement noise estimation method for an adaptive Kalman filter application and also augments external measurements in the filter to aid the position solutions, as well as uses different filters to deal with various situations. On the one hand, the adaptive Kalman filter makes use of the redundant measurement system’s difference sequence to estimate and tune noise variance instead of employing a traditional innovation sequence to avoid coupling with the state vector error. On the other hand, this method uses the external height and heading angle as auxiliary references and establishes a model for the measurement equation in the filter. In the meantime, it also changes the effective filter online based on the number of tracked satellites. These measures have increasingly enhanced the position constraints and the system observability, improved the computational efficiency and have led to a good result. Both simulated and practical experiments have been carried out, and the results demonstrate that the proposed method is effective at limiting the system errors when there are less than four visible satellites, providing a satisfactory navigation solution.

An Adaptive Low-Cost INS/GNSS Tightly-Coupled Integration Architecture Based on Redundant Measurement Noise Covariance Estimation

“…Assuming that Z 1 (k) and Z 2 (k) are independent redundant measurements of a signal Z(k) from two systems, the measurements can be modeled as [ 27 ]: where, for the measurement system i, S i (k) is the unknown measurement system error and V i (k) is zero-mean white noise at time epoch k. The first-order-self-difference (FOSD) ∆Z i and the second-order-mutual-difference (SOMD) ∆Z 12 are defined as: …”

“…In practice, because the statistical characteristics are stable over a relatively short period, a sliding window [ 27 ] can be employed to derive the autocorrelations in Equation (14). The sliding window width can be empirically set to 30~60.…”

“…Besides, altitude information is employed to augment the observation system. The ellipsoid equation is employed to serve as an auxiliary measurement equation [ 27 ]: where h is the altitude information, and are the lengths of the Earth’s semi-major and semi-minor axes respectively.…”

“…To overcome the aforementioned limitations of these existing approaches, the present work proposes a novel redundant measurement noise covariance estimation (RMNCE) approach, which employs redundant information obtained from two independent measurement systems [ 27 ] to estimate their corresponding measurement noise properties. The RMNCE approach is then applied to develop a RMNCE-based tightly-coupled (RMNCE-TC) architecture, including a new RMNCE-based satellite selection algorithm and a RMNCE-based adaptive unscented Kalman filter (RMNCE-UKF).…”

“…The RMNCE approach is then applied to develop a RMNCE-based tightly-coupled (RMNCE-TC) architecture, including a new RMNCE-based satellite selection algorithm and a RMNCE-based adaptive unscented Kalman filter (RMNCE-UKF). Additionally, the altitude aiding algorithm [ 27 , 28 , 29 ] is adopted to augment the external measurements for aiding the navigation solutions. The main contributions of this research are summarized below: A novel RMNCE approach is put forward and proved mathematically.…”