Adaptive Compensation H∞ Filter with Convergence Criterion for SINS’ Transfer Alignment

Abstract

Transfer alignment under the unknown external environment is a challenging problem in strapdown inertial navigation system (SINS). To suppress the filtering divergence when conducting transfer alignment, an adaptive compensation H∞ filter is proposed. This paper analyzes the cause of filtering divergence carefully and accordingly makes the corresponding adjustment and optimization. With the change of the unknown external environment, the robustness factor of the proposed filtering method can be dynamically adjusted. To solve the problem that the transfer alignment time is mainly determined according to experiences, the convergence criterion of the proposed filtering method is studied. The convergence criterion is based on the principles of mathematical statistics, and can automatically determine the convergence time of the filter. The aerial transfer alignment experiment illustrates that by using the convergence criterion, the alignment time can be effectively shortened and the response speed can be effectively improved.