University of Western Macedonia
Mechatronic product modularization in building safety systems, industrial and maritime applications
Abstract
dc:descriptionThe purpose of this dissertation is to contribute both theoretically, by introducing a novel algorithmic clustering approach, and practically, by proposing a suitable clustering algorithm and a decision-support framework for engineers working on complex mechatronic products. To achieve this, the dissertation evaluates the clustering results of several clustering algorithms on modularizing mechatronic products according to the structural and functional interdependencies among their components. Modularization is one of the most effective methods that industries use to drive profitability. This technique enables efficient management of complex products by dividing them into smaller clusters of components, thereby reducing the associated risks and the costs related to their design and production. This dissertation combines two distinct but complementary approaches: an unsupervised clustering study involving 175 real and simulated test products and a human-in-the-loop (HITL) evaluation using clusters generated by human experts for six real mechatronic products.Six clustering algorithms were examined in this dissertation: K-Means, Partitioning Around Medoids (PAM), Ward's method (WARD), Divisive Analysis (DIANA), Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and Clustering by Fast Search and Find of Density Peaks (DPC). The performance of the clustering algorithms is evaluated using multiple validation indices, including the Silhouette Coefficient (SC), Composed Density Between and Within Clusters (CDbw), Agglomerative Coefficient (AC), and Divisive Coefficient (DC), in conjunction with expert-based clustering results. In this dissertation, the expert-generated clusters (which serve as benchmarks), and the two-dimensional visual analyses of the clustering outcomes are employed to identify the most suitable algorithm for modularizing mechatronic products, while ensuring consistency with expert judgment in real-world applications.
Degree
thesis:*- Grantor dc:publisher
- University of Western Macedonia
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Mikrou, Ioannis
- Μικρού, Ιωάννης
Subjects
dc:subject × 10Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/60656