Vehicle Routing And Speed Optimization With Stochastic Demand And Development Trend Of Carbon Emission Control

Environmental and energy have become a focus issues in recent years.In order to reduce greenhouse gas emissions,many international agreements related to carbon emissions was formulated,which has given birth to the global carbon market.Transportation industry is a trade of high energy consumption,so it is necessary to achieve energy saving and emission reduction.Besides,in the typical vehicle routing problem,demand was usually considered to be certain.But in the real environment,demand may be random,which will affect the distribution cost.Based on this,this paper mainly studies vehicle routing optimization and speed optimization under uncertain demand.This research can help enterprises to reduce the distribution cost on the condition of meeting the needs of users and achieve the goal of energy saving and emission reduction,which has important practical significance.First of all,considering the constraints of vehicle speed,capacity and time window,this paper establishes an uncertainty model of path and speed optimization according to the relationship between fuel consumption and driving speed.And then transforms it into a certainty model by chance constrained programming.Then,two-stage algorithm are designed through model analysis: the first stage is to optimize the path suing recursive algorithm under the assumption that the network speed is known.And the second stage is to optimize the speed knowing the path.Afterwards,examples are given to verify the feasibility and effectiveness of the algorithm.What’s more,the relationship between distribution cost and customer satisfaction is also analyzed.Finally,genetic algorithm is designed because the two-stage algorithm is slow and can’t solve large-scale problems.In this paper,fmincon function and Big Integer is embedded in original genetic algorithm to optimize the speed and improve algorithm efficiency.Finally,the effectiveness and feasibility of the genetic algorithm are verified by comparing with the two-stage algorithm and solving large-scale example.