| The Capacitated Vehicle Routing Problem(CVRP)with capacity constraints is an important combinatorial optimization problem widely used in logistics transportation,scheduling optimization and other fields.With the rapid development of the economy and the high dependence on the logistics industry,the in-depth study of this issue has important theoretical significance and practical value.This thesis focuses on the following two types of extended models for CVRP:(1)The real-time logistics VRP problem with time window is proposed,and the stochastic vehicle flow constraint is added;(2)Combined with the special scenario,the thesis puts forward the problem of garbage collection with potential infection risks.In order to obtain the optimal solution that meets the requirements of the problem,the following detailed research contents are carried out:The first,considering that the travel time of vehicle distribution is not ideal due to the interference of vehicle flow,a real-time logistics VRP problem model based on the Markov model and the balance between transportation and service dual cost objectives is established,and a targeted hybrid genetic algorithm was designed to solve the problem.Firstly,by analyzing the relationship between the time attribute of data in the Solomon dataset and the order of customer delivery,a local search algorithm that can improve service quality is designed,and it is added to the standard genetic algorithm to form a hybrid genetic algorithm suitable for solving problems.Secondly,in order to verify the effectiveness of the proposed model and algorithm for real-time logistics distribution,experimental comparisons are made with the Shortest Distance Service First algorithm(SDSF),the Minimum Traffic Service First algorithm(MTSF)and Genetic Algorithm(GA)on 56 test samples.Finally,the experiment proves that the hybrid genetic algorithm can not only achieve the two optimal goals of transportation cost and service cost,but also get 100% high-quality distribution evaluation on28 samples,eliminating the penalty cost.The second,taking the sudden special events as the research background,considering the transport goods are affected by special environment,which leads to the transmission property of the virus attached to it.A model of residents’ garbage collection problems based on the infection risk model is established and design a genetic algorithm based on double optimization to solve the problem.Firstly,by analyzing the transmission characteristics of the virus,the infection risk model in this paper was established considering the two factors of time and distance.At the same time,the garbage collection problem model takes the minimum infection risk as the primary goal,and the minimum vehicle travel distance as the secondary goal,that is,the recycling work selects the recycling route with the shortest distance under the premise of ensuring the minimum infection risk.Secondly,some standard international calculation examples and simulation examples are used for experimental verification.Among them,in order to highlight the role of the infection risk model in the recycling work,the genetic algorithm based on double optimization is also used to solve the garbage collection problem with the shortest distance as the goal for experimental comparison.Finally,the experiment proves that the double optimization can improve the convergence speed of the genetic algorithm and obtain the garbage collection route plan with the lowest infection risk value.Compared with the standard VRP model,the proposed infection risk model has a positive effect on the recycling work. |
PDF Full Text Request