OBSTACLE AVOIDANCE FOR AUTONOMOUS UNDERWATER VEHICLE

• Main Author: Habibullah, Natsir
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/40401
• Institution: Institut Teknologi Bandung

Indonesia has invaluable marine potential. Therefore, Indonesia must master marine technology. One of the marine technologies that are currently developing is the Autonomous Underwater Vehicle (AUV). AUV is an underwater vehicle that can operate independently. Depending on the sensor equipment and actuators implanted in the AUV, AUV can carry out certain submarine missions. In carrying out missions on the high seas, there is a risk that AUV will collide with other marine objects, this collision can cause leakage or damage to parts of the AUV. As a result, the mission carried out by AUV can become a failure, and even worse the AUV can be lost in the middle of the ocean, even though its manufacture requires a lot of resources. Therefore, collisions must be prevented by instilling obstacle avoidance capabilities on the AUV. Before AUV can avoid obstacles, AUV must be able to find out the position of the obstacles around it. The underwater environment is a challenge in observing objects in it. Distance sensors and images that can be used in the air such as lasers or cameras cannot be used under the sea due to the characteristics of the medium in the form of water that is very different from air. One of the sensors that can be used to observe the environment underwater is a forward looking sonar (FLS). ITB is developing a Hybrid Autonomous Underwater Glider (ITB-HAUG), the vehicle is not equipped with obstacle avoidance capabilities. ITB-HAUG is equipped with FLS sensors to detect underwater obstacles. Obstacle avoidance capabilities are needed so that ITB-HAUG can carry out missions more safely. In this study, we will examine the obstacle avoidance ability for AUV which will later be used in the ITB-HAUG. The aim of this study was to design an obstacle avoidance algorithm for AUV. Avoidance of obstacles is expected to be able to anticipate the environment without prior knowledge of the existence of obstacles. To observe obstacles around the vehicle, AUV is equipped with sonar sensors, FLS. Then the algorithm is tested in mission oriented operating suite interval programming (MOOS-IvP) simulations. The selection of MOOS-IvP as a simulation base is because the MOOS-IvP can be used directly on the vehicle to carry out the actual mission. The design of obstacle avoidance algorithms is based on obstacle observation and waypoint guidance. Obstacle observation is used to determine the conditions of active obstacle avoidance and generate new waypoint points to avoid obstacles. Next, this point is passed first before the vehicle returns to its actual destination point. The test results show the waypoint guide can run according to the expected scenario. FLS sensor simulation can make obstacle observations in its scope in front of the vehicle. The obstacle avoidance design can guide the vehicle to avoid obstacles that block the vehicle path. Obstacle avoidance designs produce a shorter path and shorter duration compared to obstacle avoidance with reflexive methods.