The application of robust adaptive cubature particle filter algorithm in initial alignment of SINS

Abstract

The initial alignment error model of strap-down inertial navigation system is nonlinear under large azimuth misalignment angle condition, and it could be processed by the particle filter algorithm. In order to solve the problem that importance density function of the particle filter is difficult to select, a new robust adaptive cubature particle filter algorithm is proposed. The equivalent weight function and adaptive factor are used to allocate information, so that the effects of observation abnormality and system model abnormality on the parameter estimation are effectively inhibited. And then the estimation accuracy of heading misalignment angle is improved. By the computer simulation and experiment verification, the robust adaptive cubature particle filter is a very effective nonlinear filtering algorithm, and its filtering accuracy is higher than the accuracy of cubature particle filter.