AUTOMATIC ROUTING OF SUBMARINE ELECTRICAL CABLES USING MACHINE LEARNING

During the design process of a submarine, there are hundreds of cables that have a large diameter and a corresponding large minimum bend radius. These cables must be modeled using 3D CAD to ensure they are not bent past the minimum bend radius and to reduce the space consumed. This process is labor-intensive and sub-optimal. The objective of this thesis is to determine the feasibility of using the model-based reinforcement learning algorithm MuZero to automatically route the cables. The specific implementation of MuZero used is MuZero-General in conjunction with a custom-built Gymnasium environment. As a proof-of-concept, a 2D model successfully routed a representation of a cable from a start location to an end location after completing training. An object, a representation of an already routed cable, was added into the model. The goal was for the agent to find a path from start to finish while avoiding the obstacle. This single obstacle significantly increased the training requirements. In this model, the agent completed a route from start to finish while avoiding obstacles. However, its path was not optimal and will require additional training, which is left for future work. Overall, it was determined that it is feasible to apply reinforcement learning techniques to route cables, although challenging. In future work, the model can be expanded, making it more representative of a real-world design scenario.

Department dc:contributor.department

Mechanical and Aerospace Engineering (MAE)

Grantor dc:publisher

Monterey, CA; Naval Postgraduate School

Year dc:date.issued

2023

Author dc:creator

• Damaso, Katelyn M.

Advisor dc:contributor.advisor

• Karpenko, Mark

Statement dc:rights

• This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.

Handle dc:identifier.uri

https://hdl.handle.net/10945/72153

OAI identifier oai:identifier

oai:calhoun.nps.edu:10945/72153