题目：ChatGPT-enabled two-stage auctions for electric vehicle battery recycling

作者: Jianghong Feng（冯江洪）, Yu Ning, Zhaohua Wang, Guo Li, SuXiu Xu*

期刊: 《Transportation Research Part E-Logistics and Transportation Review》

校内级别：T2类

出版时间:   2024，Volume 183

DOI: 10.1016/j.tre.2024.103453

摘要

The primary objective of this study is to create an advanced management system that seamlessly integrates ChatGPT, an intelligent chatbot, with a cutting-edge two-stage auctions model. This integration aims to optimize the recycling process of electric vehicle batteries, ensuring utmost efficiency. To gauge the effect of incorporating ChatGPT on participants' values, we introduce technology coefficient parameters, providing a deeper understanding of its impact. In the initial phase of the auction, a sophisticated platform or an impartial third-party appraisal firm is employed. Their task is to collect vital attribute information from potential sellers regarding the waste batteries. Subsequently, this information is carefully matched with interested buyers. To determine the appropriate number of buyers both before and after the auction, we introduce two innovative matching mechanisms. We propose the implementation of a pioneering one-sided affine VCG (Vickrey-Clarke-Groves) auction, which portrays the affine externality of buyers to other buyers, along with a weighted generalized second-price auction. Additionally, we introduce a ground-breaking mechanism known as weighted multi-unit trade reduction. This inventive process allows successful bidders from the first stage to actively participate in the subsequent auction. Within this paper, we develop auction allocation rules based on the interplay between winning sellers' supply and winning buyers' demand. The outcomes of our numerical studies demonstrate the efficacy of the two-stage auction method in addressing buyers' concerns regarding battery performance. Furthermore, we have found that the utilization of ChatGPT technology proves advantageous for both participants and the platform; if buyers have a positive affine externality on other buyers, then the platform's revenue will be reduced, while social welfare will increase. In the realm of incentive compatibility, individual rationality, and budget balancing, our proposed mechanism stands unwaveringly strong, having undergone rigorous scrutiny and demonstrating commendable performance in numerous numerical experiments. Lastly, we offer practical management insights for stakeholders in the battery recycling market. These invaluable insights are derived from our hands-on experience, providing a fresh perspective for industry managers.

关键词 ：

基金资助：National Natural Science Foundation of China (NSFC 72071093) 

责任编辑：董晓玲

审    定：黄  松