A deep-sea vehicle, which possesses large delay and non-linear stepped performance, travels on the 6000m deep-ocean sediment with a random variation in its bearing capacity. For matching a high collecting efficiency, the vehicle should travel along the planned route at the onemeter precision. For traditional control this is difficult to meet. This paper presents a controller based on the fuzzy and predictive methods. The controller can identify the sediment parameter on line. It matches the traveling velocity of the vehicle and the sediment parameter, optimizes the slide rate of the vehicle and reduces the excessive disturbance on the sediment, which is useful for preventing the vehicle from losing its travel capacity because of sticking in the sediment. This will guarantee that the vehicle can travel securely. Based on the identifying results on the sediment parameter, a proper predictive respond model is chosen with respect to the sediment. Through the predictive control, the traveling speed of the vehicle is reduplicatively optimized on line. The referring trick will balance between increasing speed and the smoothness of the adjusting process. The large delay will be compensated. The overshoot, less tune and frequent adjustment are avoided or amended. The fuzzy and predictive control can fulfill the control requirements of the vehicle travel path with high precision. The result of computer simulation and the physical test proves the efficiency of the control method.


A deep-sea bed mining system consists of one or two mining ships or platforms, 5000m mineral lifting pipe, relay station, 600m flexible hose and vehicle on the seabed. The Fig. 1 shows its layout (Wang, 1994). The vehicle is looked up on the tracking target and the whole mining system goes synchronously with it at the matched velocity. Therefore, the vehicle becomes the key role in the mining system.

This content is only available via PDF.
You can access this article if you purchase or spend a download.