Distributed Model Predictive Control for cooperative floating object transport with multi-vessel systems

Autonomous technology is becoming popular by the day and has mainly been applied for the development of autonomous vehicles. Recently, this novel technology has found its way into the marine world with the development of the Autonomous Surface Vessel (ASV). ASVs have been developed specifically for the purposes of improving the safety and efficiency of waterborne transport and have grown to one of the hottest research topics. Reputable literature has demonstrated that compared to an individual ASV, greater efficiency and operational capability can be realized by groups of ASVs operating in a coordinated fashion. The technicalities of operating an ASV vary largely from those of operating a robot inland in terms of dynamics and control constraints. Further, formation tracking, i.e. connecting ASVs to an object using physical connectors; say ropes, has been shown to yield better results. This technique aims at controlling the vessels to maintain a prescribed configuration and to follow predefined trajectories. Unfortunately, collision avoidance is usually not taken into account, or only considered the conflicts among the formation mates.

Overall, noteworthy studies have shown that when the ASVs are connected to the object, changing the formation shapes to avoid obstacles is impossible. Shortfalls such as this one set the impetus for further research. On this account, researchers from the Department of Maritime and Transport Technology at Delft University of Technology: Professor Linying Chen, Dr. Hans Hopman and Dr. Rudy R. Negenborn, proposed to resolve the encountered control problem of cooperative floating object transport, i.e., utilizing a team of ASVs to transport a larger floating object, such as a large vessel, barge, or offshore platform. Their work is currently published in the research journal, Ocean Engineering.

To begin with, the researchers first described the cooperative object transport system. Subsequently, the dynamic model of an ASV and a model of towline were introduced. Here, the control strategy for object transport was proposed. The researchers then designed a multi-layer cooperative control scheme, following which an ADMM based DMPC framework was proposed to reach consensus on the following actions to be taken among the controllers. Eventually, the scenarios in which the proposed cooperative system moves a large vessel sailing inbound the Port of Rotterdam were simulated to show the effectiveness of the proposed method.

In their approach, the cooperative transport problem was divided into three sub-problems, trajectory tracking of the object, control allocation, and formation tracking of the ASVs. Overall, the results obtained showed that the proposed cooperative system could transport the floating object along a predefined trajectory and avoid potential static and dynamic obstacles.

In summary, the study focused on the cooperative floating object transport, i.e., a group of ASVs coordinate their actions to transport floating objects. Professor Linying Chen and colleagues proposed a formation-based cooperative object transport system with a multi-layer control structure. In a statement to Advances in Engineering, Professor Linying Chen said that their approach would ultimately lead to methods that could become useful for moving large vessels, barges, and off-shore platforms in future ports where both human-operated and autonomous vessels exist.