| Summary: | 碩士 === 國立中山大學 === 海下科技暨應用海洋物理研究所 === 100 === In the development of underwater vehicles, it is necessary to conduct performance test in the water tank. However, different factors need to be considered depending on different cases. The purpose of this paper is to construct a simulation platform in the air to study the scenarios like side current or constant height profiling. Although these tests are difficult to be performed in the water tank, we can get some solutions from observing the dynamics of simulation platform. The simulation platform we used consists of three links to constraint the motion in a polar coordinate system. It carries a wireless micro-camera, and two DC motor-driven propellers. At the end of the distal link and metal disk is attached on the other side of the pivot of the last link to provide counter-weight which can simulate different status of the "buoyancy" of the platform. The encoder which is uses to trace the motion of the simulation platform, is mounted at each join between two links. The control program has two parts: servo control of propellers and target tracking. In order to approach to the real-time searching, we derived image with gray scale instead of color form to increase image refreshing rate during the tracking process. For the current experiment, the target is at dot generated by an LED. The location of the bright dot is detected by a histogram-based threshold, and the actual location is further refined with intensity-weighted algorithm. The offset between of the target and the center of the image is used as the feedback to command the propellers to drive the platform. The goal is to keep the target at the center of the image as close as possible. A linear PD control (proportional - derivative) is implemented to drive the propellers. Preliminary experiments show that the simulation platform can track a target with about 15 frames per second refreshing rate under the condition that the target does not move too fast and vanishes in the image. When ROV with laser scanner syetem, seafloor away from ROV''s accuracy is necessary. In this eassay, we use tracking angle and tracking bright dot to qualify and quantify the influence of buoyancy and propeller on the altitude control in different cases.
|
|---|
