An Investigation into Interdependencies in Electrical Machines Involving Deformable Materials: A Model-Based Multi-Objective Framework

The increasing popularity of the green industrial revolution has led to a rise, in the production and use of Electric Vehicles (EVs) and their associated parts, like motors, generators and transformers [1]. This surge was driven by a growing emphasis on sustainability and the need to shift towards alternatives especially in transportation [2]. However meeting the growing demand for these parts while maintaining production standards presents a challenge. To address this challenge it is important to improve manufacturing processes by prioritizing early detection of faults to reduce End of Line (EoL) tests and ensure efficient production. The efficiency and cost control of Electrical Machines (EM) greatly depend on the interactions, between components and subsystems. The inclusion of deformable materials such as copper wire introduces complexity owing to their nature. This study suggests various frameworks to improve efficiency by minimising the need for tests and effectively handle deformable materials during the manufacturing process. In this research, a framework that utilises Discrete Event Simulation (DES) was presented to investigate the interrelationships in EM manufacturing specifically focusing on the coil winding process. Through experiments, connections between winding speed, wire thickness, bobbin shape and variations in resistance were discovered. These findings highlight the importance of control features. Additionally, this study introduced an improved DES framework that combines the original DES model with a Supervised Machine Learning (SML) algorithm via Knowledge Distillation (KD). This integration significantly reduced simulation times while still maintaining accuracy. Lastly, a new approach was introduced with the aim to optimise the linear coil winding process by considering multiple objectives. The goal was to minimise production costs and decrease faults by examining the connections between these objectives. Advanced techniques like the NSGA-II algorithm were utilised to find a balance between faults and production costs resulting in enhancements, in operational efficiency and cost reduction.