INS/ADS/GPS组合导航高灵敏度故障检测和识别方法

The chi-square fault detection method cannot precisely locate the specific dimension of the subsystem variables, and sensitively detect soft fault. An INS/ADS/GPS high-sensitivity fault detection and identification algorithm is proposed. The proposed algorithm establishes two sequential fault detection models of INS/GPS, INS/ADS by introducing sequential filter. To solve the problem of low sensitivity of soft fault detection, forgetting-sequential probability ratio test(F-SPRT) based on fading memory factor is proposed. Combining F-SPRT with sequential chi-square detection, the high-sensitivity fault detection and identification architecture is realized. The simulation results show that under the condition that the false alarm rate is 0.1%, compared with the traditional sequential probability ratio test (SPRT) method, the sensitivity of the proposed algorithm is increased by 2 times respectively, and fault mode can be identified effectively.