Guidance and motion control logic for Marine Autonomous Surface Ships

The increasing maritime traffic and a major awareness regarding environmental pollution lead to a greater emphasis on the efficiency of operations and their costs. Improvements can be achieved with the progressive integration of automation tools to relieve humans from repetitive and complex tasks that might lead them to error or to non-optimal solutions. The promises are to enhance both human and environmental safety while also holding the potential for increased efficiency and reduced operational costs. The introduction of automated systems occurs at different levels with applications spanning from decision support systems to fully autonomous ships, defined as Marine Autonomous Surface Ships (MASS). Both academic institutions and industrial companies carried out various studies on this topic over the past decades, leading to the design and testing of pioneering prototypes of fully autonomous vessels. The above-mentioned studies focus on a broad spectrum of areas that contribute to the design of a decision support or autonomous system. The topics include situation awareness systems for obtaining information on the surrounding environment, systems for route planning or for following the desired path, systems for avoiding collisions with fixed or moving objects or for automatic mooring, and more. Different methodologies can be defined for each of the above mentioned tasks and can be organised through the well-known schematisation in the three independent modules of the Guidance, Navigation, and Control (GNC). For the motivation explained above, the research carried out within this thesis fits into this context, leading to the guidance and control modules. The main objective is the development of systems for planning and controlling motions through the entire speed operational range of surface vessels. Different kind of vessels leading to model-scale to full scale vessels are adopted for this study. With respect to the state of the art, this work would like to propose an integrated motion control scenario suitable for all the MASS operation conditions, from berthing, track and station keeping, to path following, and target tracking operations. Controlling the motion at low or high speed involves different design approaches that can be schematised into 2 or 3 degree-of-freedom (DOF) controllers due to the effects of the actuators in the sway motion at high surge velocities. The outlined 2 DOF motion control scenario, used for high speed, aims to follow a moving object controlling the desired heading and speed. The adopted guidance law is the target tracking one. It consists of reaching and following a target of which it is only possible to know the instantaneous position and velocity, hence when the future motion is unknown. In this context, the standard Line-Of-Sight, Pure Pursuit, and Constant Bearing laws are modified and deeply tested. The controller layout comprises an autopilot and a speed pilot, working together to establish the desired guidance setpoint, and the synthesis leads to provide the stability of the closed-loop together with some performance characteristics. The presented 3 DOF motion control scenario desires to maintain position or to follow a path at low speed. The chosen guidances enable smooth setting of the desired setpoint using reference models or to define a trajectory to be tracked for track keeping. Several control layouts are investigated, composed of a controller together with the force and thrust allocations. Controllers consist of both feedback contributions like PID or its variations and a feedforward component computed through a model-based approach or disturbance reconstruction. The allocation problem is deeply investigated and solved through both a simplified approach and optimization techniques. Both 2 and 3 DOF motion control scenarios are applied to real test cases in both model and full scale, each characterised by different main dimensions, manoeuvrability capability, and propulsion system configurations. Each test case is described and their characteristics are shown. Specific key performance indicators, running from integral metrics to yearly operability and emission indexes, are selected for each motion control scenario to evaluate the effectiveness of the proposed GNC modules. The proposed motion control scenarios are extensively tested using non-linear dynamic simulators representing the MASS test cases. Based on the results, the defined key performance indicators are used to assess the effectiveness of the proposed approaches across a spectrum of speeds and environmental conditions. In conclusion, this thesis contributes to the field of autonomous marine vessels by providing motion control strategies suitable for the entire speed range.