Canonical and Non-Canonical Approaches of the Discrete Teaching Learning Based Optimization

Manufacturing plays a crucial role in robust economies, and effective optimization of manufacturing processes ensures competitive viability. Multi-hole drilling is a fundamental process in manufacturing, especially in the mass production of boilerplates, food processing separators, drum and trammel screens, and printed circuit boards, where optimizing drill tool paths is crucial for cost competitiveness. Multi-hole drill tool path sequencing is often framed as Traveling Salesman Problems, known for their NP-hard complexity. Researchers employed evolutionary algorithms to tackle the challenges associated with these NP-hard complexities. The recently proposed Discrete Teaching Learning Based Optimization (DTLBO) can address the intricacies of multi-hole drill tool path sequencing. This study proposes the use of DTLBO for multi-hole drill tool path sequencing optimization and highlights the critical distinctions between Canonical and Non-Canonical approaches of the DTLBO. Further, it assesses their performances through a comparative analysis using test problems, investigating the merits and demerits of these approaches. The findings of this investigation provide a foundation for refining the algorithms and improving their practical effectiveness, which will be pursued as future research. Notably, no prior research in the literature has undertaken such a comprehensive comparison.