Research Of Second Distribution Of Product Oil Problem Based On Adaptive Large Neighborhood Search Algorithm

With the continuous development of society,the demand for product oil in industries such as automobiles,machinery,industry,and agriculture is increasing.Product oil refers to the oil products produced through the production and processing of crude oil.In daily life,our common gasoline,kerosene,diesel,ethanol gasoline,and biodiesel are all finished oil.In the product oil logistics system,highway distribution costs account for as much as 60% to 70% of the total logistics costs.Reasonable optimization of the delivery route of oil tankers can improve the efficiency of petroleum product distribution,the guarantee of petroleum supply,and the satisfaction of consumer groups,which is an important link for petroleum companies to further reduce costs,increase efficiency,and improve competitiveness.This thesis studies the secondary distribution of product oil.The process of transporting product oil from multiple oil depots to various gas stations through multiple types of tank trucks(single tank vehicles,multi tank vehicles)is a real life logistics planning issue.The vehicle routing problem has developed different variants to describe the problem more realistically.Based on existing research,this thesis aims to minimize the total vehicle usage cost,transportation cost,and penalty cost of the time window,and establishes a vehicle optimal scheduling model that includes six VRP variants: hybrid model,split loading,delivery with time windows,multi trip delivery,multi depot delivery,and demand splitting.The purpose is to more conform to and be closer to the actual situation.A heuristic algorithm is designed to solve the model,and a savings algorithm is used to solve the initial distribution plan.Subsequently,a hybrid algorithm combining adaptive large neighborhood search algorithm and simulated annealing algorithm is used to optimize the initial distribution plan to obtain the final distribution plan for product oil.Finally,based on the data obtained from field research in Dalian City,different scale calculation examples are constructed.The effectiveness of the model and algorithm proposed in this thesis to solve the secondary distribution problem of product oil is verified through small,medium,and large-scale examples.It is possible to derive the following rules from the experimental data: 1)The two cabin fleet has certain advantages in saving time window penalty costs and vehicle operating costs,while the single cabin fleet has obvious advantages in saving vehicle operating costs.Using heterogeneous fleets can fully leverage the advantages of single cabin and multi cabin vehicles,resulting in a reduction in overall delivery costs.Compared to single cabin homogeneous fleets,using heterogeneous fleets resulted in average cost savings of 13.03%,7.93%,and 4.69% for small-scale,medium-sized,and large-scale fleets,respectively.Compared to the two cabin homogeneous fleet,the average cost savings are 10.27%,14.63%,and 10.35%;2)Multi-trip distribution,as opposed to single-trip distribution,can effectively reduce the number of vehicles used and lower total costs by around 46.44%.These rules can serve as a decision-making guide for businesses that distribute refined oil,helping them to cut costs and increase productivity.