Vehicle And Cargo Matching For Steel Logistics

In the process of the informatization transformation of the steel logistics industry,many stubborn problems in the industry have not been effectively solved,such as truck overload,overdue orders,backlog of cargoes and other issues.The reason is that the spatial and temporal distribution of bulk logistics capacity resources is not uniform.The traditional vehicle and cargo matching methods tend to take the maximum of each truck's load weight as the optimization goal,without taking the priority of orders,warehouse cargoes due date,the completion rate of vehicle,cargo matching and other attributes into consideration,all of which have a critical impact on the profit of steel logistics platform.Therefore,under the premise of limited transportation capacity resources,this thesis aims to establish appropriate vehicle and cargo matching mechanism for different transportation periods and platform bottlenecks,so as to improve the overall revenue of the platform.Aiming at the challenges of high-priority cargo backlog and low completion rate of vehicle and cargo matching in the steel logistics industry,this thesis provides an appropriate vehicle and cargo matching decision-making mechanism for the steel logistics platform.First of all,through the analysis of a large number of real data from steel logistics,the important rules related to the process of cargoes distribution are extracted.On this basis,according to serious economic loss caused by high priority cargo backlog problem,we present a method based on the target of maximizing the portion of high priority shipment of cargoes.Specifically speaking,the method uses a improved genetic algorithm to packaging the cargoes as loading plans,then builds graph structure in each time window,finally utilizes the breadth-first search algorithm to achieve an available and near optimal matching results.On the other hand,in view of aggravated capacity loss caused by the cargoes backlog problem,the thesis put forward a matching method based on the matching degree of trucks with the goal of maximizing the weight of cargo.The method first uses dynamic programming algorithm to realize the packaging process of cargo to loading plan,then utilizes greedy algorithm to complete the combination of the cargoes in remaining orders.Under the technical supports which are the adaptive window and bipartite graph matching created by vehicle and loading plan,the method realizes real-time vehicle cargoes matching decision,and increases of long-term future earnings for steel logistics platform.The main work of this thesis includes the following aspects:· This thesis analyses the mechanism of the vehicle and cargo matching module in the logistics platform,and provides a new idea for the construction of the vehicle and cargo matching module in the logistics platform by combining the analysis results of big data and the actual situation of the vehicle and cargo matching.This thesis summarizes the current development status of vehicle and cargo matching,sorts out the existing methods of vehicle and cargo matching,classifies the principle of various vehicle and cargo matching methods at home and abroad,and points out that they are not suitable for the scene in this thesis.· For high-priority cargoes backlog caused serious economic loss,this thesis is proposing a cargo loading plan decision framework based on the target of high-priority cargo delivery maximization.The framework puts forward a improved genetic algorithm to packaging the cargoes as loading plans,then builds graph structure in each time window,finally utilizes the breadth-first search algorithm to achieve an available and near optimal matching results in real time scenario.· In view of the low utilization rate of transportation resources,this thesis presents a vehicle and cargo matching framework which aims to maximize shipment weight and minimize the number of tail orders.Firstly,the framework packages the cargoes as loading plans,and then uses bipartite graph matching algorithm to complete the matching of vehicles and load plans.Finally,according to the training of the historical data,the framework designs adaptive window to get the appropriate time window cutting,and completes the matching process between vehicle and cargoes in real-time scenario.