A Fast Gradual Fault Detection Method for Underwater Integrated Navigation Systems

Abstract

Gradual fault detection is always an important issue in integrated navigation systems, and the gradual fault is the most difficult fault to detect. To detect gradual faults in a timely and precise manner in integrated navigation systems, the statistical concepts of the normalised residual mean and the sum of absolute residuals are introduced according to the characteristics of gradual system failure in this paper. The applicability of the improved residual χ 2 detection method is discussed. Then, the gradual fault detection program based on the improved residual χ 2 detection method is designed with the criterion of normalised residual mean and the sum of absolute residual. The simulation results and vehicle tests show that: 1) The residual of the failed sub-system can be calculated accurately with the improved residual χ 2 detection method, which has strong applicability in gradual fault detection; 2) The gradual fault can be detected in a short time by using the normalised residual mean and the sum of absolute residual.2. χ 2 detection method. 3. Gradual fault detection. 4. Mean of residual.5. Sum of absolute residual.

The Influence of Satellite Configuration and Fault Duration Time on the Performance of Fault Detection in GNSS/INS Integration

“…When the false alarm rate is defined as α, according to the Neyman–Pearson criterion, the threshold Td can be worked out through solving the equation P{sans-serifΛ¯k>Td}=α [10]. The fault detection criteria is:{sans-serifΛ¯k>Tdfault occurssans-serifΛ¯k≤Tdno fault occurs…”

“…Until now, many mature fault detection methods have been proposed, which can be classified into three categories, i.e., hardware redundancy methods, analytical redundancy methods, and nonanalytical redundancy methods [6,8]. The hardware redundancy configuration usually exceeds the minimum necessary and increase the cost of the navigation equipment [9,10]. The nonanalytical redundancy methods essentially are data-driven methods based on machine learning [6].…”

“…Using recursive filtering and the residual chi-square test, an integrity and quality control procedure called detection, identification, and adaptation (DIA) was investigated by Teunissen [19]. Compared to the state chi-square detection method, the residual chi-square detection method can detect the abrupt fault in time with a small amount of computation, but it does not work well in gradual fault detection [10]. In order to improve the performance of chi-square fault detection method for gradual fault, many improved methods had been proposed.…”

Adaptive Fault Isolation and System Reconfiguration Method for GNSS/INS Integration

“…A new rate detector algorithm based on AIME was developed in [12] and the test results show that it has better detection performance than AIME and MSS for gradual fault. A fast gradual fault detection method based on improved residual chi-square detection method was proposed for underwater integrated navigation systems [13]. Besides, fault detection based on artificial intelligence has attracted more attention and been widely studied.…”

“…However, the above researches mainly aimed at how to detect the fault accurately and fast, the fault isolation and system reconfiguration, which is another key issue of the fault detection and tolerance, was seldom considered. In multisensor integration, usually the faulty sub-filter is isolated and the measurement update process is accomplished by fusing the results of the other normal sub-filters [13]. In GNSS/INS integration, isolating the faulty subsystem will force the integrated system into the pure INS model, which will result in the decrease of the filtering precision [16].…”