Analyzing vehicle path optimization using an improved genetic algorithm in the presence of stochastic perturbation...

Area of Science:

• Operations Research
• Logistics and Supply Chain Management
• Computational Intelligence

Background:

• Vehicle path optimization is crucial for efficient logistics and distribution.
• Stochastic perturbations, time window variations, and load capacity constraints complicate route planning.
• Carbon tax mechanisms add an economic layer to distribution strategies.

Purpose of the Study:

• To develop a perturbation scheduling model for logistics and distribution incorporating a carbon tax.
• To propose an enhanced Genetic Algorithm (GA) with a Gaussian Matrix Mutation (GMM) operator for improved optimization.
• To address time window variations and load capacity constraints in vehicle path optimization.

Main Methods:

• Established a perturbation scheduling model for logistics considering stochasticity and carbon tax.
• Developed an enhanced Genetic Algorithm (GA) utilizing a Gaussian Matrix Mutation (GMM) operator.
• Constructed a Gaussian probability matrix based on site positional order for gene mutation.
• Employed a roulette-wheel-selection method for guided population evolution.

Main Results:

• The enhanced GA demonstrated a 50-60% increase in average convergence speed compared to the classical GA.
• Algorithm execution time was reduced by 48% with the proposed method.
• Solution accuracy was maintained within a 1% difference.
• Experimental simulations validated the effectiveness of the GMM-based GA.

Conclusions:

• The enhanced GA effectively optimizes vehicle paths under complex constraints and carbon tax policies.
• The GMM operator successfully balances population diversity and convergence speed.
• The proposed model offers a significant improvement in efficiency for logistics and distribution operations.