Multi-strategy Quantum Differential Evolution Algorithm And Its Application In Vehicle Route Problem

Vehicle Route Problem(VRP)is a kind of point-line topology problem in real scenarios,which is of great significance to the modernization and upgrading of Logistics System and the construction of Intelligent Logistics System.With the rapid increase of the demand of the logistics industry and the increasingly complex domestic road conditions,the vehicle routing problem presents complex features,such as large-scale,multi-objective,and strong constraints,which makes the traditional accurate algorithm and heuristic algorithm difficult to solve efficiently and accurately.Therefore,it has important theoretical significance and practical application value to an efficient and robust new intelligent optimization algorithm to effectively solve the complex vehicle routing problem.Differential Evolution Algorithm(DE)is taken as the main research object,and the positive effect of the mutation strategy on convergence accuracy and speed is fully used to propose a differential evolution algorithm based on neighborhood mutation and oppositionbased learning,namely NBOLDE.In NBOLDE,the new parameters and weight factors are introduced into the neighborhood to design a new neighborhood mutation strategy(DE/neighbor-to-neighbor/1)based on the DE/current-to-best/1 mutation strategy.Then,the opposition-based learning is employed to optimize the initial population to accelerate convergence,improve efficiency,and enhance stability.Based on the discreteness and constraints of VRP,the strategy in NBOLDE is adjusted.The Gravitational Search Algorithm(GSA)and the local search strategy are introduced to effectively solve the small-scale and lowdimensional vehicle route problems.In order to overcome the shortcomings of DE in solving large-scale problems,such as low solution efficiency,insufficient diversity in the late search stage,slow convergence speed,and easy to fall into stagnation,the quantum computing characteristics and cooperation of QEA and the divide and conquer idea of CCEA are combined.A differential evolution algorithm based on hybrid mutation strategy and cooperative co-evolution framework(HMCFQDE)is proposed.In HMCFQDE,a quantum differential co-evolution framework is designed,in which subpopulations divided for independent solutions.The hybrid mutation strategy is employed to improve the search efficiency,and the quantum rotation is employed to enhance the diversity of the population.To effectively realize the large-scale vehicle route problems,the strategy in HMCFQDE is adjusted,the sub-population framework and the local search strategy is fully used to design the differential mutation and quantum rotation methods for effectively realizing large-scale vehicle route planning.Aiming at the problems of large computational complexity and high time complexity in solving high-dimensional large-scale vehicle routing problems,the cloud computing with virtualization,high reliability and strong generality is introduced.A solving method of highdimensional large-scale vehicle route problem based on cloud computing and HMCFQDE is proposed to effectively reduce the running time and improve the solution efficiency.