Integrity monitoring of lowcost GPS-aided-INS systems

As unmanned systems become more and more important, reliability and integrity issues become definite, specially when being implemented with low-cost (or sometimes are referred to as COTS) sensors while being designed to operate in harsh environments. As a result, fault (or failure) detection and identification (FDI) is a must, and is a crucial requirement for designing unmanned vehicles. In this work, we investigate the utilization of two FDI techniques, the chi² gating function and the the multiple model adaptive estimation (MMAE). Chi-squared FDI systems are computationally very inexpensive, have good fault detection ability and require no a priori knowledge on system dynamics. However, they are sensitive to filter tuning and fail to detect faults when the filter converges to them rather than rejecting them. Model-based approaches provide outstanding FDI ability. However, they are computationally demanding, require a priori knowledge on system model, sensitive to mismodelling errors, have finite convergence time and compromise filter optimality under no-failure conditions. A new FDI fusion algorithm is proposed, which guarantees integrity and does not affect optimality under no-failure conditions. Simulated results are presented to highlight performance characteristics of both FDI system implementations.