Climate issues have become more and more important now. Although global governments have made some progress, we are still facing the truth that the prospect of international cooperation is not clear at present. Due to the limitations of the Integrated assessment models (IAMs) model, it is difficult to simulate the dynamic negotiation process. Therefore, using deep learning to build a new agents based model (ABM) might can provide new theoretical support for climate negotiations. Building on the RICE-N model, this work proposed an approach to climate negotiations based on existing trade groups. Simulation results show that the scheme has a good prospect.

翻译：当前气候议题日益重要。尽管全球各国政府已取得一定进展，但国际合作的预期前景仍不明朗。由于综合评估模型（IAMs）的局限性，无法有效模拟动态谈判进程。据此，利用深度学习构建新型智能体模型（ABM）或能为气候谈判提供新的理论支撑。本研究在RICE-N模型基础上，提出基于现有贸易集团的气候谈判路径。仿真结果表明该方案具备良好发展前景。