Marine Vehicles with Streamers for Geotechnical Surveys: Modeling, Positioning, and Control*

In the classical set-up adopted for marine geotechnical surveys, one or more acoustic sources carried by a vessel emit signals that penetrate the seafioor, reflect or refract off geological features, and travel back to multiple acoustic receivers, or hydrophones, placed along cables (streamers) towed by the same vessel. Processing the data obtained by the hydrophones yields information about the physical properties of the Earth's crust. In this set-up, the acoustic sources and receivers are coupled and the geometry of the latter is fixed. The EC WiMUST project aims to contribute to the development of a new breed of reflection seismology systems whereby the source and the receivers are physically decoupled and free to change their spatial configurations. The project puts forward a novel concept where shorter streamers are towed by autonomous marine vehicles (AMVs) that are required to maintain a possibly time-varying formation. The paper describes the first steps taken by 1ST, in cooperation with project partners, towards the development of the networked navigation and control systems of the WiMUST vehicles. We first address the problem of modeling a coupled streamer-vehicle system and propose a solution based on Kane's method to obtain a dynamic simulation model. A brief description of the on-going work on model parameter identification is given. We then focus on the cooperative navigation problem and propose a simplified version of a nonlinear least squares approach previously described in the literature. Finally, we address the problem of cooperative control of a fleet of streamer-vehicle systems. We give an overview of cooperative path-following and target/trajectory tracking techniques and discuss possible adaptations to align them with the technical requirements imposed by the WiMUST concept. Preliminary simulation and experimental results are shown.