PH-SHOWOA: Parallel hybrid SHO-WOA for VRPSPDTW

Area of Science:

• Operations Research
• Metaheuristics
• Combinatorial Optimization

Background:

• The Vehicle Routing Problem with Simultaneous Pickup and Delivery and Time Windows (VRPSPDTW) is a complex logistical challenge.
• Existing metaheuristics face limitations in balancing exploration and exploitation for VRPSPDTW.

Purpose of the Study:

• To develop a novel parallel hybrid metaheuristic for solving the VRPSPDTW.
• To enhance solution quality and computational efficiency for VRPSPDTW.

Main Methods:

• Integration of Spotted Hyena Optimizer (SHO) for diversification and Whale Optimization Algorithm (WOA) for intensification.
• Implementation of an adaptive probability control mechanism for dynamic search behavior regulation.
• Incorporation of simulated annealing, local search, and population diversification strategies for robustness.
• Parallel implementation for concurrent solution updates and local refinements.

Main Results:

• PH-SHOWOA achieved an average reduction in total travel distance of over 10% compared to standalone SHO and WOA.
• The proposed method demonstrated competitive and often superior performance against advanced algorithms like Co-GA, MA-FIRD, and ACO-DR.
• PH-SHOWOA achieved optimal solutions on several benchmark instances and performed well in centralized-demand scenarios.
• Statistical tests confirmed the effectiveness and robustness of the PH-SHOWOA algorithm.

Conclusions:

• The parallel hybrid metaheuristic PH-SHOWOA effectively addresses the VRPSPDTW.
• PH-SHOWOA offers a robust and efficient approach, outperforming existing methods in terms of solution quality and computational efficiency.