Automated Maneuvering in Confined Waters using Parameter Space Model and Model-based Control

The paper discusses methods to increase the level of automation in ship handling towards a possible autonomous operation. The focus is on maneuvering situations in confined waters within the velocity range between dynamic positioning and transiting. While performant automation solutions exist for specialized vessels, standard ships are operated manually in maneuvering situations. In this context, one challenge is to adapt a model for controller design of maneuvering vessels. It is a cumbersome task to parameterize the common hydrodynamical oriented models, especially for maneuvering standard ships. Therefore, a more experimental approach has been chosen to decrease the complexity of the model structure. In that way, the applied motion model is highly abstracted and has a minimal number of parameters which are mapped in parameter spaces. For motion control, a cascaded structure integrating a velocity and a maneuver control system has been designed. The low-level part consists of a model-based feedforward control applying the parameter space model implicitly. Further a simple decentralized multi-variable feedback controller is used. Here, a robust approach has been applied for controller parameterization by assigning a specific parameter space to each defined operation range. The methods are verified and validated with two demonstrators. Firstly a passenger vessel is used in a ship handling simulator and secondly real world experiments are performed applying an unmanned surface vehicle. The objective of these trials is automated maneuvering in the port of Rostock. Copyright (C) 2020 The Authors.