Adaptive fuzzy fault-tolerant voting mechanism based on EKF
Jinxue Sui, Zhen Hua, Li Yang, Yu‐Xin Tian, Yan Zhang · 2008
In this paper, the adaptive fuzzy fault-tolerant voting mechanism has been proposed based on EKF, which can solve the bottleneck problem of voting when there are many fault-tolerant gyroscopes floating point input signals. Through connecting EKF in the original fuzzy voting mechanism, the adaptive fuzzy fault-tolerant voting mechanism has prevented the flaw that the fuzzy voting is unable to judge when the signal has a smaller deviation. Moreover, by designing another logical controller in EKF and combining EKF filter and the fuzzy logic method regulation measured residual and covariance matrix, we have achieved the real-time adjustment of the Q covariance and R covariance matrix of the EKF, which obtains better convergence value and avoids EKF causing the divergence phenomenon because of the mode abnormal or the information sloughing in the solution process. After simulation of four sets of gyroscopes that represent four sets of inertial navigation system, this adaptive fuzzy voting mechanism can decide the correct output, and this output may solve the question of the measured noise and the gyro error.