Multi-objective optimization using improved NSGA-II for integrated process planning and scheduling problems in a...

Area of Science:

• Manufacturing Engineering
• Operations Research
• Industrial Engineering

Background:

• Large-size valve production involves complex, small-series manufacturing with diverse parts.
• Simultaneous production of multiple valve models and sizes in a single workshop presents significant scheduling challenges.
• The integrated process planning and scheduling (IPPS) problem in this context is a large-scale, NP-hard combinatorial optimization challenge.

Purpose of the Study:

• To develop and validate an effective methodology for solving the IPPS problem in large-size valve manufacturing.
• To improve production efficiency by minimizing makespan and reducing manufacturing costs.
• To bridge the gap between theoretical IPPS research and practical industrial applications.

Main Methods:

• An improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) was developed for IPPS.
• A two-section encoding and inserting greedy decoding method were employed within the algorithm.
• Dynamic population update strategies and adaptive mutation techniques were utilized to enhance solution quality and diversity.

Main Results:

• The proposed NSGA-II algorithm successfully obtained satisfactory solutions for makespan and manufacturing costs.
• The methodology demonstrated effectiveness in a real-world case study at Yuanda Valve Company.
• The implemented approach considered realistic manufacturing constraints, improving practical applicability.

Conclusions:

• The improved NSGA-II algorithm offers a viable solution for the complex IPPS problem in large-size valve manufacturing.
• The integration of dynamic population updates and adaptive mutation enhances algorithm performance.
• The study successfully demonstrates the practical application of advanced optimization techniques in a real manufacturing environment.