Multi-objective AGV scheduling in an automatic sorting system of an unmanned (intelligent) warehouse by using two...

Area of Science:

• Logistics and Supply Chain Management
• Operations Research
• Artificial Intelligence in Logistics

Background:

• Automated Guided Vehicles (AGVs) are crucial for unmanned operations in intelligent warehouses, enhancing efficiency and competitiveness.
• Optimizing AGV task scheduling is complex due to factors like charging requirements and variable speeds.
• Existing scheduling methods may not fully address multi-objective optimization for AGVs.

Purpose of the Study:

• To develop a multi-objective mathematical model for AGV task scheduling.
• To optimize AGV operations by considering charging tasks and variable speeds.
• To minimize makespan, the number of AGVs utilized, and electricity consumption.

Main Methods:

• Development of a multi-objective mathematical model for AGV task scheduling.
• Integration and comparison of two adaptive genetic algorithms (AGA) and a multi-adaptive genetic algorithm (MAGA).
• Numerical experiments to evaluate algorithm performance and model validity.

Main Results:

• The multi-adaptive genetic algorithm (MAGA) demonstrated superior performance compared to the adaptive genetic algorithms (AGA).
• Optimization resulted in significant improvements, with objective values changing by approximately 30%.
• The developed model and MAGA effectively addressed the complexities of AGV task scheduling.

Conclusions:

• The proposed multi-objective model and MAGA are rational and valid for optimizing AGV task scheduling.
• MAGA offers a robust solution for enhancing warehouse efficiency and reducing operational costs.
• The study validates the effectiveness of advanced algorithms in solving complex logistics optimization problems.