해양 탐사를 위한 새로운 AUV 플랫폼의 설계 및 제어

Abstract

This dissertation presents a design, control, and implementation of a new autonomous underwater vehicle (AUV) platform for ocean exploration, which is designed as a torpedo shape with small and light enough to be handled easily by one person. According to a unique ducted propeller and rudder located at the aft, the AUV can perform horizontal motion. It can also control pitch angle and depth motion by an inside mass shifter mechanism (MSM) which changes the vehicle center of gravity. In addition, hardware and software architectures of the control system are addressed and the functions of all parts are described.

Navigation system design is a critical step in determining position, course, and distance traveled of a vehicle. Hence, in this dissertation, a navigation algorithm based on discrete Kalman filter is developed to estimate the states and to compensate accumulative errors. Moreover, experiments were carried out to verify the performance of the developed algorithm using the developed navigation system of AHRS, GPS-INS, and DVL-INS.

Based on the nonlinear six degree of freedom AUV model and multivariable sliding mode control method, the controllers of heading and depth are designed. After that, with the developed navigation system and suggested controllers, a number of experiments were performed and their results are compared with the simulation outputs.

In practical operation, the AUV needs to avoid obstacles such as reefs and seawalls in the littoral environment. For this, an algorithm of obstacle avoidance was devised. Also, a modeling of the sonar sensor that gathers information of the ocean environments and the forward obstacle for safe navigation was established.

So this dissertation focuses on examining the behavior and control system required for the AUV to maneuver over obstacles in the vertical and horizontal planes. Through the developed obstacle avoidance algorithm, the AUV has ability to use range and bearing data received from sonar modeling to determine if that return constitutes a threat along its desired path and further navigate around the threat before regaining its original path. In addition, it is difficult and unsafe to maneuver AUV to avoid obstacles with constant surge velocity, the velocity controller with output of the propeller thrust is also considered.

Finally, to validate the developed obstacle avoidance algorithm and controllers, simulations in vertical and horizontal planes were performed and their results were described.