Gravitational Apparent Motion-Based SINS Self-Alignment Method for Underwater Vehicles

Jingchun Li,Wei Gao,Ya Zhang

DOI: https://doi.org/10.1109/tvt.2018.2876469

IF: 6.8

2018-01-01

IEEE Transactions on Vehicular Technology

Abstract:To solve the self-alignment problem of strapdown inertial navigation system (SINS) for underwater vehicles, a novel gravitational apparent motion (GAM)-based method is proposed. Different from conventional GAM methods, the proposed GAM method can complete SINS self-alignment under swaying conditions without using the a priori local latitude information. First, we determine the gravity vector in the earth frame and the local latitude by using the gradient descent optimization and certain geometry constraints. Then, the self-alignment process is formulated as an optimization-based alignment quaternion determination problem by constructing an objective function with the estimated gravity vector. We employ gradient descent optimization to achieve the least square solution of the objective function. Thus, the attitude quaternion can be determined according to the quaternion product chain rule. The simulation and experiments results demonstrate the proposed GAM method without using the local latitude achieves an alignment accuracy close to conventional GAM methods during the coarse alignment process.