Abstract
The paper is dedicated to the overview of the autonomous ship onboard control system functionality. The paper is dedicated to autonomous maritime navigation, i.e. to planning a safe and optimal route for an autonomous ship amid the absence of crew on board and to handling an autonomous ship when underway, ensuring seaworthiness and route correctness. Moreover, the paper describes communications of several autonomous ships and autonomous and ordinary ships to infer a collective decision on how to pass safely through a particular area. Since it is expected that autonomous ships are going to be equipped with dozens of sensors and detectors, the paper describes remote monitoring of an autonomous ship when underway. Since an autonomous ship highly depends on its onboard control system, the paper also pays attention to the robustness of the system. The paper suggests evolutionary computations as a solution for the route planning problem because this approach enables to perform multicriteria optimization of a set of solutions simultaneously. For autonomous ship handling and seaworthiness control, the paper suggests machine learning techniques because these techniques can solve problems in case of uncertainty and the environmental mutability. For inter-ship communications, the paper suggests distributed consensus algorithms, widely used in parallel and distributed computation systems. To ensure the onboard control system robustness, the paper suggests the actor approach that represents that the entire software system consists of a set of elementary agents communicating in the distributed computational environment.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have