A crowdsourcing logistics solution based on digital twin and four-party evolutionary game

Abstract

With the development of mobile Internet and the popularity of public online consumption, the scale expansion of the e-commerce industry has driven the continuous growth of logistics business, but the increase in order volume has brought pressure and challenges to the end of logistics and distribution, and the emergence of crowdsourcing logistics It provides a new path for alleviating the current logistics and transportation dilemma. The Digital Twin (DT) can virtualize and learn the data of the physical space, and introduce DT into the crowdsourcing logistics. It can iteratively update the crowdsourcing logistics participant strategy by constructing a virtual space, so as to change the corresponding strategy in time. Considering the situation of crowdsourcing logistics workers signing contracts with platforms and colluding with platforms, this paper constructs a four-party evolutionary game model of temporary workers, contract workers, blockchain-based crowdsourcing platforms and task requester, and analyzes the evolution using replication dynamics method stabilization strategy. In the virtual scene of DT, multi-agent reinforcement learning is used to predict the evolution result of the current strategy, and a reward and punishment strategy is given to prevent workers from free-riding and platform false reporting. The simulation results show that the analyzed evolutionary stability strategy can make the crowdsourcing logistics system run continuously and healthily, and the participants can adjust the strategy correctly according to the prediction results of virtual crowdsourcing logistics.