Hybrid Quantum-behaved Particle Swarm Optimization For Disruption Management In Logistics Distribution Problem

Due to the imbalance between supply and demand in the agricultural products industry,there is often a problem of unsalable overprice of products.The new selling model,probabilistic selling,can effectively alleviate the problems caused by imbalance between supply and demand by providing customers with homogenous agricultural product packages online.At the same time,the logistics distribution combined with the probabilistic sales characteristics and the disruption management in the distribution process are also a new logistics distribution scenario.The main research work of this paper is as follows:(1)Construct a model of vehicle route problem with time window and propose a solution algorithm.Based on the previous research on vehicle route problem,a mathematical model of vehicle route problem with time window and capacity constraint is constructed.Based on the standard particle swarm optimization algorithm,adding crossover operator and genetic operator to propose a hybrid quantum-behaved particle swarm optimization.As the main algorithm for solving the model,the algorithm compares with the results of classical intelligent algorithms such as standard particle swarm algorithm,tabu search algorithm,simulated annealing algorithm,genetic algorithm,and evaluates the advantages and disadvantages of the algorithm.(2)The multi-stage and multi-objective disruption management model is established and solved when two kinds of disruption events occur simultaneously in the logistics distribution system.In this paper,the disruption event is extended to the time window change and the customer position change simultaneously.Firstly,according to the disruption event,it is judged whether the initial scheme is feasible.When it is not feasible,the multi-stage multi-objective disruption management model is established.Considering the different participants,such as the customer,operator and dispatcher in the system,the first stage takes the delivery completion rate as the objective function,the second stage takes the route length,vehicle amounts and the number of new road as the objective function.At the same time,taking the priority of customers into account,the hybrid quantum particle swarm algorithm is used to obtain the adjustment plan.(3)The probabilistic selling is chosen as the selling model and is quantitatively analyzed to anti-disruption ability by experiments.In order to make up for the lack of traditional selling,this paper adopt the probabilistic selling model.This selling model is characterized by providing customers with a homogenous product package,so an inter-route scheduling strategy can be used when an disruption event occurs.By comparing the adjustment plan of the traditional selling and the probabilistic selling,we can find the advantage of probabilistic selling model.This paper improve the particle swarm optimization algorithm which was used for continuous problems to apply for the discrete vehicle route problem.At the same time,the crossover operator and mutation operator are introduced,and the hybrid quantum particle swarm optimization algorithm is proposed to improve the convergence and global search ability of the algorithm.Based on the initial plan and the idea of disruption management,we consider the situation that two disruption events occur simultaneously,build a multi-objective and multi-stage interference management model,enrich the actual logistics distribution system,and provide theoretical support for vehicle real-time scheduling and route plan decision.