Study On The Filter Method Of INS In Transfer Alignment

Transfer alignment is a kind of initial alignment technique which use the main INS with higher accuracy to calibrate the slave INS. It is a key technique for the new generation of rapid reaction and motor launch weapons.Although main INS with high accuracy, its' error affects transfer alignment precision of slave INS. In this paper, however, the solution error and the kalman filter estimation bias are simulated for the given transfer alignment model in the case of external velocity and position detection error. The analysis results indicate that the kalman filter estimation bias resulted from external detection error can be neglect in the case of main INS with high accuracy and noise of slave INS is white noise. However, kalman filter is bad choice when systemnoise isn't white noise.Also, we apply to H_∞ filter methods. H_∞ filter requires the disturbance noise is energy bounded other than known accurate statistic. H_∞ filter has robustness both in uncertain noise and in uncertain system parameter, and it canbe completed in the conditions mentioned above.The method of designing the H_∞ filter based on LMI is presented in thispaper. It is convenient to design a filter in the case of system uncertainty denoted by multi-cell. It is indicated in the INS transfer alignment that the error angles can be convergent by the robust H_∞ state estimator based on the LMI for the systemwith a uncertainty. It provides rapid alignment as well as robustness against uncertainty for INS transfer alignment.