A hybrid multi-level ant colony optimization framework for integrated production scheduling and vehicle routing

Area of Science:

• Operations Research
• Supply Chain Management
• Computational Optimization

Background:

• Time-sensitive supply chains require efficient operational-level scheduling.
• Integrating production and distribution is complex, especially with vehicle routing and time windows.
• Existing methods may struggle with large-scale, time-sensitive scheduling problems.

Purpose of the Study:

• To develop a hybrid multi-level optimization (HMLO) framework for integrated production and distribution scheduling.
• To address parallel machine scheduling and vehicle routing problem with time windows (VRP-TW) with flexible departure times.
• To optimize distribution costs using the ant colony system (ACS) within the HMLO framework.

Main Methods:

• A hybrid multi-level optimization (HMLO) framework was developed.
• The problem was decomposed into parallel machine scheduling and distribution scheduling phases.
• Ant Colony System (ACS) metaheuristics and heuristics were integrated for optimization.

Main Results:

• The HMLO framework yielded optimal solutions for small-scale instances.
• The framework outperformed existing methods, including LINGO software, for medium and large-scale instances.
• An average improvement of 5-30% in solution quality was observed for large-scale instances compared to LINGO.

Conclusions:

• The proposed HMLO framework is effective for integrated production and distribution scheduling.
• The method offers superior convergence and solution quality for complex supply chain problems.
• This approach provides significant performance improvements over traditional software solutions.