Research On Multi-Objective Joint Optimization Of Loading Problem And Vehicle Routing Problem

The precision of logistics services and efficient business operation become the core competitiveness of e-commerce enterprises with the electronic commerce coming into the era of internet+. As an important theoretical foundation of logistics operations, vehicle routing problem is always one of the core issues in the field of logistics distribution, and it is very important to improve the efficiency of logistics distribution. So vehicle routing problem has got high attentions of academic and business since it arised. However, people put forward higher requirements for vehicles management to meet the precision of the logistics service, together with the closely related loading problem. Today's society is facing social environment problems such as road congestion, vehicle emissions. With the cost of personnel and vehicle maintenance increasing, reducing the number of vehicles and improving loading efficiency are implemented as important means to reduce operating costs and improve resource utilization. Thus the logistics distribution management combining vehicle routing problem and loading problem is becoming a new hotspot in the field of logistics distribution at present and in the future. For this reason, this dissertation considers the joint problem of two NP-hard problems which are loading optimization and vehicle routing problem, and investigates and explors the multi-objective modeling and solving on the joint problem. The main researches are listed as follows:(1) Modeling and solving for the multi-objective joint optimization of loading problem and capacitated vehicle routing problem (LCVRPMO)This dissertation applied the theories and methods of multi-objective dicision into capacitated vehicle routing problem (CVRP), and established the multi-objective optimization mathematical model for the basic joint problem of loading and CVRP. Focusing on part of Pareto solutions, a somewhat succinct model called multi-objective CVRP facing defferent target preferences (MOCVRPFDTP) was proposed. There are three preference structures in it, which are joint optimization preference, absolute minimum number of vehicles preference and path optimization preference. The dissertation built an efficient algorithm architecture including the corresponding specific algorithms to solve the three preference structures respectively. In the implement part, the closely related problems of path optimization and packing optizimation were also researched, and severl algorithms were designed.(2) The multi-objective joint optimization of 3D-loading and CVRP with different preference.Aiming at the demand of more general 3-dimensional boxes, multi-objective joint optimization of 3 dimentional loading and CVRP was proposed, and the corresponding mathematical model (3LCVRPMO) was established. The dissertation put forward an algorithm architecture including different strategies of multiple phases/layers along with specific algorithms to slove the path optimization preference of 3LCVRPMO. The existing relevant algorithms about 3-dimensional packing were studied and improved to be more suited for joint optimization problem. This dissertation also put forward the algorithm ideas for the other two subproblems of 3LCVRPMO which are the joint optimization preference and the absolute minimum number of vehicles preference.(3) Research on 3LCVRPMO with time window constraintsThis dissertation established the mathematical model of 3LCVRPTWMO for the problem of 3LCVRPMO with time window. An effective algorithm architecture together with relevant specific algorithms was proposed to solve it. The solving of CVRP with time window (CVRPTW) has mainly focused on the path optimization objective before. This dissertation also gave the other corresponding results such as the number of distribution vehicles and waiting time, which made CVRPTW better realize multi-objective optimization of various resources and services.This proposed joint multi-objective model and relevant algorithms are effective, which was verified through a series of experiments and data analysis. These studies enrich the basic theory of CVRP and packing problem, and provide beneficial exploration and more solutions for the precision service of practical logistics business.