Probability-scenario guided adaptive large neighborhood search for stochastic order allocation

Area of Science:

• Operations Research
• Environmental Science
• Logistics Management

Background:

• Multimodal freight logistics faces challenges in assigning uncertain orders under carbon emission constraints.
• Balancing transportation costs, carbon penalties, and expected profit is crucial for sustainable operations.

Purpose of the Study:

• To develop a high-dimensional stochastic optimization model for maximizing expected profit in freight transportation.
• To address carbon emission constraints within logistics planning.
• To propose an efficient intelligent optimization approach for solving complex logistics problems.

Main Methods:

• Development of a high-dimensional stochastic optimization model.
• Integration of a probability-guided adaptive large neighborhood search with scenario generation.
• Prioritization of influential order combinations and key scenarios for computational efficiency.

Main Results:

• The proposed intelligent optimization approach demonstrated significant improvements in solution quality and computational speed compared to conventional methods.
• Objective performance improved by over 10% on average.
• Computational time was reduced by more than 80% compared to a baseline random sampling algorithm.

Conclusions:

• The developed approach offers a robust and stable solution for high-dimensional stochastic logistics optimization under environmental constraints.
• It provides practical insights for enhancing sustainable logistics planning.
• The method effectively balances economic objectives with environmental considerations in freight transportation.