Kalman–Hatch dual‐filter integrating global navigation satellite system/inertial navigation system/on‐board diagnostics/altimeter for precise positioning in urban canyons

Abstract

For improved positioning in urban canyons, this study proposes an efficient dual‐filter method integrating multi‐constellation global navigation satellite system (GNSS), an inertial navigation system (INS), a barometric altimeter and an on‐board diagnostics (OBD) module. The proposed method consists of a position‐domain (PD) Hatch filter and velocity Kalman filter. The Hatch filter is operated as the main positioning filter and the Kalman filter is operated as a sub‐filter to aid the main filter for the occasional GNSS outages. The Hatch filter combines barometric altimeter measurements with GNSS pseudorange and carrier phase measurements. The Kalman filter integrates the inertial measurement unit (IMU) measurements with GNSS Doppler shift and OBD speed measurements. A semi‐simulation method is applied for the accurate performance evaluation of the proposed method compared with the several conventional methods under deficient satellite visibility, multipath errors and cycle slips caused by urban canyons. By the test results, it is demonstrated that the proposed method can bound the positioning errors effectively by utilising low‐cost all‐weather sensors. The RMSEs of the proposed Kalman–Hatch dual‐filter algorithm were shown as 0.05, 0.19 and 0.11 m, respectively, and the maximum errors were shown as 0.16, 0.26 and 0.40 m, respectively, in the east, north and upward directions.

A tightly‐coupled compressed‐state constraint Kalman Filter for integrated visual‐inertial‐Global Navigation Satellite System navigation in GNSS‐Degraded environments

“…When utilising the carrier phase measurements, cycle slips and multipath errors need to be detected carefully. For this purpose, the Fault Detection Exclusion (FDE) algorithm [6,32] is applied. In this manner, the proposed CSCKF-based VIG system provides globally consistent and robust pose estimates even in the GNSSdegraded environments.…”

“…In order to overcome the time‐increasing INS errors, the integration of the Global Navigation Satellite System (GNSS) and INS has been an effective alternative [4, 5]. By utilising GNSS receivers, the integrated system is not confined by estimating local [6] DoF poses but it can also provide globally referenced position estimates. In recent years, the development of multi‐GNSS boosted the progress of INS/GNSS systems.…”

“…[7] demonstrated that a multi‐GNSS system provides better performance than a Global Positioning System (GPS)‐only system for INS augmentation. However, the performance of GNSS is strongly influenced by abnormal signal conditions occurred in urban canyons and tunnels [6]. For instance, if the time length of GNSS signal outages is long, the positioning accuracy provided by any INS/GNSS system will be degraded due to the rapid accumulation of INS errors.…”

Multipath/NLOS Detection Based on K-Means Clustering for GNSS/INS Tightly Coupled System in Urban Areas

“…However, satellite signals are frequently blocked or even lose lock in complex urban scenarios, which cannot guarantee the effectiveness of positioning [ 3 ]. Inertial Navigation System (INS) has the advantages of strong autonomy and strong anti-interference and can obtain short-term, high-precision navigation and positioning results [ 4 ]. However, INS errors accumulate over time, and long-term independent solution can result in reduced accuracy or even divergence [ 5 ].…”

Adaptive Doppler‐smoothed‐code bilateral kernel regression method for single‐frequency BeiDou receiver

“…In the range domain, it is a trend to reform the CSC parameters for a better position solution, such as smoothing window width [7,8], alternative observation selection [9] and ionosphere mathematical method [10], which sometimes may lead to position divergence [11]. In the position domain, phase-connected filter, Hatch filter with Kalmantype gain and covariance estimation method are proposed [12][13][14] to cope with the structure shortage in the CSC smoothing process, which can solve the weakness of the range-domain method well, but always require strict parameter estimation. Besides, due to external correction data, the CSC method performs well and is widely used in augmentation systems since many error sources ignored by the CSC method, especially the ionospheric delay, can be eliminated.…”