Teaching learning-based optimization for cost-oriented hybrid flow shop scheduling / Wasif Ullah

Faculty of Manufacturing and Mechatronic Engineering Technology

Thesis (Master of Science) -- Universiti Malaysia Pahang – 2025

Includes bibliographical references

Production scheduling is a strategic process of organizing the execution of jobs on available resources to optimize certain optimization objectives. One of the important scheduling problem is the cost-oriented hybrid flow shop (CHFS) scheduling problem, which involves optimizing the scheduling of jobs across multiple stages to minimize the scheduling-related costs. Despite the significance of cost optimization in CHFS scheduling, there is a lack of comprehensive studies that address all major cost components using efficient optimization algorithms. For this purpose, a comprehensive cost model was developed that includes four cost elements: labor costs, machine energy consumption costs, preventive maintenance costs, and late penalty costs. Then a Greedy-assisted Teaching Learning-Based Optimization (GTLBO) algorithm was proposed to optimize the developed CHFS model. In GTLBO algorithm the initialization process was hybridized with Greedy algorithm, where 10% of initial population was generated by Greedy algorithm, while 90% randomly generated as usual. Afterward, a computational experiment was conducted to evaluate the performance of GTLBO algorithm by choosing certain comparative algorithms. The experiment utilized a dataset of 12 benchmark test problems defined by Carlier and Neron and was conducted using MATLAB version 2022b. The Wilcoxon rank-sum test confirmed statistically significant improvements (p value < 0.05) in most scenarios. The experiment and Wilcoxon test revealed that GTLBO outperformed other algorithms in optimizing the CHFS problems. For the practical validation of the CHFS model and the GTLBO algorithm, two real-case study problems were analyzed by comparing the optimized schedule costs obtained using GTLBO with the costs of the original schedules. In case study 1, the proposed method reduced the total scheduling cost from RM 53,314.00 to RM 50,854.88, achieving a cost reduction of approximately 4.6%. In case study 2, the cost was reduced from RM 214.35 to RM 208.92, resulting in a 2.5% improvement. This research can help the manufacturers with HFS scheduling setups struggling to minimize their production expenses. Future research directions include establishing multi-objective CHFS scheduling models and developing hybridized optimization algorithms to optimize these models.