Application and Efficiency Analysis of Genetic Algorithm in Multi-Vehicle Path Optimisation Models Under Time Window Constraints

In modern logistics and supply chain management, the multi-vehicle routing problem (MDVRPTW) is a highly complex optimisation problem involving multiple distributed warehouses and scheduling of different types of vehicles. The core of this problem lies in how to minimise the total cost while satisfying specific time window requirements. This paper provides insights into the application of genetic algorithms and their efficiency in solving the multi-vehicle path optimisation problem with time window constraints. Genetic algorithms search through the solution space by mimicking natural selection and genetic mechanisms using operations such as selection, crossover, and mutation to find an approximate global optimal solution. The article begins with a review of the history of research on the vehicle path problem and its variants, emphasizing the complexity of the time window and multi-vehicle garage problems. It then details the implementation of genetic algorithms, including path-based coding strategies, elite retention combined with roulette wheel selection methods, partial mapping and sequential crossover methods, as well as swapping and reversing mutation operations. Through a series of designed experiments, this study demonstrates the efficiency and stability of the genetic algorithm in dealing with large-scale MDVRPTW problems. The experimental results show that the genetic algorithm can not only effectively optimise vehicle scheduling and path planning, but also significantly reduce the operating cost. Finally, the article discusses the performance of the genetic algorithm, analyses its advantages and limitations in practical applications, and provides a reference for future research directions.