Optimizing port yard storage at Dar es Salaam Port: a hybrid approach combining mixed-integer linear programming and genetic algorithms

Introduction: Efficient port yard storage is a critical challenge at Dar es Salaam Port, where limited space, diverse cargo types, and strict regulatory requirements complicate operations. Traditional optimization methods often fail to address these complexities adequately. This study presents a hybrid approach that combines Mixed-Integer Linear Programming (MILP) with Genetic Algorithms (GA) to optimize storage operations, aiming to improve efficiency and adaptability. Methods: The methodology involves a two-phase optimization strategy. In the first phase, a robust MILP model is developed using the PuLP library in Python to determine optimal cargo allocations based on volume, compatibility, and equipment constraints. The second phase refines these allocations using a Genetic Algorithm, facilitated by the DEAP library, to explore and optimize solutions further. This hybrid approach is designed to handle the complexities of real-world port operations by leveraging the strengths of both MILP and GA. Results: The MILP model successfully allocated cargos to storage locations, optimizing space utilization and reducing handling costs. The largest storage area, Location L5, was utilized up to 95% of its capacity, reflecting an efficient allocation strategy. The Genetic Algorithm further enhanced the initial solution, increasing overall utility by 3% and providing multiple near-optimal configurations.