This paper represents the experimental investigation on the position keeping control of a scaled ROV model in a basin. The dynamics of ROV motion is modeled by a simple linear equation of motion. The 3-dimensional position and heading of ROV are measured by an ultrasonic positioning system using a digital signal processor. The acoustic positioning system uses linear frequency-modulated pulses and the matched filtering technique to reduce the effect of ocean environmental noise and to improve the range resolution of the acoustic positioning system. Two control systems using a classical PID controller and a linear quadratic Gaussian controller are tested. The effects of modeling errors and measurement errors from the acoustic position measuring system on the performance of the designed control system are studied. Dynamic position keeping tests are performed for 4 degrees of freedom motion which are surge, sway, heave and yaw.
This paper describes the design and performance of a position control system for a 500 meter class ocean survey ROV model developed in Korea Research Institute of Ships and Ocean Eng. (KRISO). There are many papers on the design of ROV position control systems. Sliding mode techniques deal directly with those problems effectively (Yoerger, Newman & Slotine, 1986, 1991), (Suzuki, 1991). Adaptive controls with recursive system identification techniques are frequently used in uncertain models (Yuh, 1990), and in models with disturbances (Goheen, Jefferys & Broome, 1987). The system parameters for each operating condition are found with either PMM tests or system identification experiments. Goheen & Jefferys (1989, 1991) have described general methodologies which can be used for the performance prediction and autopilot design. Many investigators who work in simulation study, however, have ignored sensor noises encountered in real systems.
