Optimal transportation service trading mechanism in e-commerce logistics

Abstract

E-commerce logistics is an internet-enabled logistics value chain designed to offer competitive logistics service including transportation management, warehousing management, and freight consolidation. However, current logistics in e-commerce industry has become a bottleneck, and fails to keep pace with soaring orders. This thesis proposes the concept of E-commerce transportation service trading (ETST) and develop both optimized algorithms and theoretical methods to coordinate the trading relationship between e-commerce platform/retailer and logistics service providers (LSPs). ETST is the problem of determining and optimizing of the matching issue between the sets of e-commerce orders generated from e-commerce platform and the sets of transport vehicles offered by LSPs respectively. Four typical scenarios are investigated in this research. The first scenario establishes a framework for formulating ETST problem with the consideration of its characteristics. A cost minimization model is developed to optimize the assignment of B2B online orders and vehicles. A full-factorial simulation experiment is proposed to investigate the effects of factors on the performance indexes such as the total logistics cost per unit, the type of vehicle required, and the number of vehicle invested. The results show that the vehicle cost and order characteristics (ratio of weight and volume) significantly impact the vehicle assignment in terms of the vehicle type and vehicle quantity. The second scenario investigates a less-than-truckload carrier collaboration (LTLC) decision making problem in the e-commerce logistics network. This scenario first proposes the concept of an e-commerce logistics trading system with collaborative decisions. A collaborative transportation planning (CTP) model is introduced to maximize the total profit without reducing the individual profit of the carriers with information sharing. This scenario then proposes a stochastic plant pollinator algorithm and conducts extensive computational experiments. The results show that the higher degree of cooperation will help to achieve more benefits for carriers. The third scenario conducts a game theoretic analysis on the situation where two logistics service providers (LSPs) compete in an e-commerce logistics market with respect to the order quantity decision and service level decision on a particular logistics service product. The analytical results and the equilibrium analysis of the e-commerce logistics system are investigated. After that, the equilibrium results are compared among the centralized system, Cournot game and Stackelberg game. A revenue sharing contract are designed according to the decentralized model. A numerical study is conducted to illustrate the impact of parameters on the optimal decision variables. The fourth scenario addresses a vehicle routing problem with simultaneous pickup and delivery with time windows from multiple depots (MVRPSPDTW) over a time horizon in the B2C e-commerce logistics system. This scenario considers an e-commerce logistics system with multi-period, which consisting customers, logistics service providers (LSPs), suppliers and a decision-making platform. A mixed integer linear programming model is developed for the problem and tested on small scale instances. To handle more realistic large-scale problems, a differential evolutionary algorithm (DE) and parallel differential evolutionary algorithm (Par-DE) metaheuristic is proposed. The computation experiment is conducted on the real data from practical scenario and comparative result is demonstrated.