Development Of ROV With Underwater Structure Detection And Research On Navigation And Dynamic Positioning

Remotely Operated Vehicle(ROV) is an efficient tool to complete underwater tasks, which can perform assignments such as detection, maintenance and inspection. Many underwater structures require detection and determination periodically along with the continuous exploiting on ocean resources. And this situation demands that a ROV should not only float under water, but being equipped with the ability to determinate while climbing along sidewalls. Therefore, it is imperative to develop a new-type ROV which have abilities of floating and climbing whereas can inspect and cleans. Combined with industry-studyresearched prospective joint project of Jiangsu Province, this paper participates in designing and developing a prototype of Model-Converted ROV(MC-ROV) which is able to switch between floating model and climbing model freely, as well as researches of navigation and dynamic positioning. The details are listed as follows:This paper illustrates the electrical system design of the MC-ROV. It is divided into two blocks: electronic tank and motor driving tank. The former one contains visual unit, motion converted unit, alarm and protection unit, sensor unit and main control unit, while the latter one is mainly related to motor drive and its protection and cooling.The upper console monitoring and control system of MC-ROV is designed, which is comprised of computer, control console and joystick that function as Human Machine Interface(HMI), display and control. HMI is developed based on VB, including state presentation, function control, preserving data, 3D posture and alarm. In view of underwater motion of MC-ROV,a MEMS-based integrated navigation system is designed, which covers one 3-axes accelerator, one 3-axes gyroscope, one 3-axes compass and one depth gauge, while the chip of STMF4 is utilized as a controller. At the same time, navigation system is solved.Deep study of MEMS navigation algorithms on the basis of complementary filter and Unscented Kalman filter(UKF) is launched, which achieves good experimental results in a pool. Meanwhile, researching gradient descent method in the compensation of quaternion updating and the Complementary Ensemble Empirical Mode Decomposition(CEEMD) algorithm in the noise reduction of gyroscopes and accelerators separately. Besides, according to the dynamic coefficients calculated by data simulations, dynamic model of MC-ROV is built to research the control of dynamic positioning near water surface provided that MC-ROV encounters with wave disturbance. And a new control algorithm based on Non-singular Terminal Sliding Model(NTSM) and UKF.Many underwater tests demonstrate that MC-ROV can implement pretty good motions such as forward and back, spin, setting depth, setting heading angle, presenting positive navigation and control performances that can accomplish expected requirements. And simulation results prove the efficiencies of gradient descent method, CEEMD and NTSM.