초공동 시험체 초기 개략정렬 및 오차분석

Abstract

This paper presents an initial coarse alignment (ICA) algorithm for a strapdown inertial navigation system (SDINS) of the prototype of a supercavitating vehicle. Initial attitude information is essential for an SDINS system to calculate the state variables. Due to the limitations of the target vehicle of this study, an inertial navigation system (INS) should be developed without a depth or velocity sensor. Without these sensors, the divergence of position and velocity is inevitable. Initial attitude information becomes more critical for minimizing the divergence of state errors. Accordingly, three types of initial coarse alignment algorithms are proposed: inertial measurement unit (IMU)-based ICA (IMU-ICA), launch tube attitude-based ICA (LT-attitude ICA), and inclinometer-based ICA (inclinometer-ICA). Monte Carlo simulation is conducted to evaluate the performance of proposed ICA algorithms. A supercavitating vehicle launch scenario is used to describe the real test environment. The SDINS algorithm, which only uses IMU, is presented. Error analysis predicts the error characteristics of proposed ICA algorithms and SDINS. Depth and attitude errors at the end of the simulation are used as a performance index. The performance of the proposed ICA algorithms using the proposed sensor specification is validated and compared. Furthermore, the required sensor specification to satisfy the desired SDINS performance is calculated based on the simulation result to compare the performance of ICA algorithms regardless of sensor specification.