Research On Multi-objective Low Carbon Vehicle Routing Optimization And Algorithm

With the rapid development of the social economy,the distribution demand of the modern logistics industry has grown rapidly,which has also caused urban environmental problems to become increasingly prominent,especially the air pollution problem led by haze,which has continued to arouse widespread concern in the society.The logistics transportation industry is one of the important sources of greenhouse gases such as carbon dioxide,and the carbon emissions generated by road freight transportation account for an important proportion in logistics transportation.How to rationally plan the vehicle distribution route on the basis of meeting the distribution requirements of small batches and multiple batches,and effectively reduce the energy consumption and carbon emissions generated during the distribution process,has gradually become a hot spot in the research of vehicle distribution routes.Based on the classic vehicle routing problem(VRP),considering the vehicle fuel consumption and carbon emission model,time-dependent road network and other issues,In-depth study of the multi-objective low-carbon vehicle routing problem under static network and time-dependent network,And the main research content includes the following aspects:(1)On the basis of the existing research work,firstly,the existing research theories and practices of vehicle routing problem,low-carbon vehicle routing problem,multi-objective vehicle routing problem and time-dependent vehicle routing problem are systematically described.From the three parts of optimization objectives,model constraints and algorithm design,the shortcomings of the existing research are summarized and analyzed,and the multi-objective low-carbon vehicle routing problem is proposed,which has a strong practical significance.(2)From the perspective of energy saving and emission reduction,first of all,the fuel consumption and carbon emission models from the macro and micro dimensions is compared,and determine use the comprehensive fuel consumption model as the fuel consumption calculation method.Secondly,based on the existing research results,the VRP problem is classified,and the basic vehicle routing problem and the model with time window vehicle routing problem are introduced.Finally,the commonly used precise algorithms,heuristic algorithms and meta-heuristic algorithms for vehicle routing problems is summarized(3)In view of the actual background of energy conservation and emission reduction in logistics distribution enterprises,a carbon emission calculation method that comprehensively considers the driving speed,load and mileage of the vehicle is proposed.The triangular distribution is used to quantify the speed change of the vehicle during the travel process.A multi-objective low-carbon vehicle routing problems with the lowest total cost of the system and minimum vehicle turnaround time is established.At the same time,in order to improve theaccuracy and convergence speed of NSGA-II algorithm in solving this kind of problems,the ideas of co-evolution and information interaction in the emerging multi-factor optimization algorithm are used to improve NSGA-II,and an enhanced non-dominated sorting is proposed Genetic algorithm(ENSGA-II).The results of the calculation examples show that the constructed multi-objective low-carbon vehicle routing model under static network can comprehensively consider the transportation interests of different logistics distribution participants,and can provide effective support for decision makers.The proposed algorithm has better search performance and solution effectiveness.(4)Because the path optimization under the static road network is an idealized optimization model,although it has certain practical significance,it is not enough to efficiently guide the actual production tasks.To this end,considering the dynamics of the road network traffic information,by analyzing the discrete travel speed,a number of time zones are divided according to the speed change,so that the driving speed of the vehicle is continuously changing,and the shortcomings of the existing discrete travel time representation methods are summarized,Discrete travel time is continuously processed to construct a continuous travel time function,that is,the travel speed of the road segment and the travel time of the road segment are time-dependent.You can call different travel speeds and calculate different travel times according to the time zone to which the departure time belongs.Non-linear time window satisfaction function to characterize customers’ psychological changes,and then calculate carbon emissions and time window satisfaction.The model can reflect the uncertainty of actual traffic information,the calculation process is more detailed,and the obtained optimization results are more in line with the actual transportation background Build a time-dependent multi-target low-carbon vehicle path model under the network,and use the ENSGA-II algorithm to solve the model,providing an effective basis for the decision-making plan of the decision maker.Sensitivity analysis of time-dependent attributes shows that the different time zones experienced by transportation tasks result in different economics and timeliness.The effects of energy saving and emission reduction are also different,making path planning more flexible.