Overcoming the settling time challenge of initial alignment of strapdown inertial navigation system by a novel solution

Abstract

This study investigates two fundamental challenges in the initial alignment of strapdown inertial navigation systems and presents creative solutions to overcome these challenges. The first challenge is to determine the settling time in different operating conditions such as stationary, quasistationary, in-motion, etc. The proposed solution to this challenge is to adjust the settling time of the initial alignment adaptively by evaluating the initial attitude matrix computed using the vector observations instead of allocating a fixed interval time for settling time. The second challenge is the existence of uncertainties in the vector observations, which adversely affect the initial alignment performance. The proposed solution to this challenge is to evaluate the quality of the vector observations used in the initial alignment by a real-time algorithm. Appropriate weights are assigned to vector observations instead of allocating a uniform weight to all vectors, depending on the vector observation's quality. The capability and performance of the proposed solutions are evaluated using different experimental scenarios such as three-axis rate table and vehicle tests.