Integrity of an integrated GPS/INS system in the presence of slowly growing errors. Part II: analysis

Abstract

This is the second of two consecutive papers (Part II) in this journal on the monitoring of the integrity of integrated GPS/INS systems. Part I established that the worst class of error for an integrated system in terms of failure detection performance is that of slowly growing errors (SGEs). It was also concluded that among the integration architectures, the tightly coupled architecture provides the best mechanism to tackle SGEs due to its simpler structure and accessibility to the relevant measurements. In a subsequent comparison of the existing integrity algorithms, the multiple solution separation (MSS) and autonomous integrity monitoring by extrapolation (AIME) methods were selected for further analysis because of their representative characteristics of the current integrity algorithms. This paper carries out a detailed investigation of the capability of the MSS and AIME algorithms to deal with the SGEs, using a realistic simulation platform and limited real data analysis. Results show that although it is possible for the existing algorithms to detect SGEs, the time it takes to detect them varies inversely to the rate of growth of the SGE. A new algorithm is developed to deal with this based on the concept of rate detection. Simulation results show that the new algorithm can detect SGEs earlier than the current algorithms. This is validated by preliminary tests with real data.

High accuracy tightly-coupled integrity monitoring algorithm for map-matching

“…Therefore, a non-perfect Gaussian error distribution needs to be considered and the HPL needs to be properly calculated so as to satisfy the integrity risk requirement. In such a case, the measurement fault and the measurement bias may become correlated with each other (Blanch et al, 2007;Ober 2003). The correlation may increase the test statistic and decrease the position error.…”

“…To compensate this deficiency, GPS has to be integrated with other sensors, such as DeadReckoning (DR) and Geographic Information System (GIS) based road map to provide horizontal positioning supplement and a Digital Elevation Model (DEM) to reduce the error in the vertical direction. Integration of data from various sources is primarily carried out in two modes (Bhatti 2007): loosely-coupled and tightly-coupled. It has been proven that a tightly-coupled system provides advantages over a loosely-coupled system, especially under the conditions of less than four satellites in-view or in the case of poor satellite geometry even if the number of visible satellites is more than four (Bhatti 2007;Quddus and Zheng 2011).…”

“…Integration of data from various sources is primarily carried out in two modes (Bhatti 2007): loosely-coupled and tightly-coupled. It has been proven that a tightly-coupled system provides advantages over a loosely-coupled system, especially under the conditions of less than four satellites in-view or in the case of poor satellite geometry even if the number of visible satellites is more than four (Bhatti 2007;Quddus and Zheng 2011). This kind of scenarios is more frequent in urban operational environments.…”

“…It is found that the effect of correlation could induce the difficulty in satisfying the integrity risk requirement in terms of slow ramp errors (Li et al, 2011). The AIME based HPL could provide the edge in dealing with the effect of slow ramp errors (Bhatti 2007).…”

Extreme value theory-based integrity monitoring of global navigation satellite systems

“…The former uses only the current measurements while the latter uses both current and historical data, together with a priori assumptions on dynamics. The major drawback of filtering methods is their insensitivity to slowly growing errors (Bhatti, 2008). The snapshot scheme is the most widespread method for RAIM due to its faster response to sudden failures compared to the filtering scheme.…”