A New Approach to Solve the Vehicle Routing Problem: The Perspective of the Genetic Algorithm Combined with Adaptive Large Neighbour Search

Abstract

With the rapid development of the modern economy, people's demand for logistics distribution services is increasing daily, and commodity distribution has become an indispensable part of social logistics activities. Vehicle routing problem (VRP), as an essential process in the distribution industry, has attracted extensive research from scholars in the fields of logistics optimization. Based on the existing algorithms, this paper proposes an improved adaptive large neighbourhood search (ALNS) algorithm combined with a genetic algorithm and simulated annealing algorithm (GA-ALNS-SA). Under the constraints of the number of vehicles, load and the maximum route distance, the fitness function is set based on the total route distance. A genetic algorithm iteratively generates the initial solution of improved adaptive large neighbourhood search. Then the optimal global solution is worked out by roulette, simulated annealing and other methods, through which the traditional operator design and selection strategy are improved. The conclusion scheme is visualized, and the corresponding solution results are analyzed, which provides an effective programme formulation method for implementing the vehicle routing problem.