센서 모델 기반 수중트랙로봇의 수중항법 시스템 연구

Abstract

In this study, an underwater navigation algorithm was developed to apply the underwater navigation system to the underwater track robot. Generally, underwater navigation uses a Doppler Velocity Log(DVL) to measure the velocity of underwater vehicles. However, undersea platforms, such as underwater track robots, cannot use DVL due to the distance limitations of sensor operation. As a result, Dead Reckoning(DR) navigation is inevitably used, and which results in severe errors in attitude and position values over long periods of platform operation. To overcome this problem, we developed an underwater navigation system composed of coupled Inertial Navigation System(INS) composed Ultra Short Base Line(USBL) and additional track information. The INS sensors were modeled using the mathematical model of the accelerometer, the gyroscope, the magnetometer. Before the experiment, computer simulations were performed to analyze the expected sensor values for specific track missions in unexpected situations. Based on this, we developed an underwater navigation algorithm for a prototype underwater track robot which we developed at the lab and confirmed the effectiveness of the navigation algorithm through experiments. For the prototype underwater track robot, we developed the navigation system, the electric hardware, the control system, and the operating system. Finally, we applied the developed INS and the underwater navigation algorithm to the platform and verified a good performance through real sea experiments.