"Industry 4.0" aims to build a highly versatile, individualized digital production model for goods and services. The carbon emission (CE) issue needs to be addressed by changing from centralized control to decentralized and enhanced control. Based on a solid CE monitoring, reporting, and verification system, it is necessary to study future power system CE dynamics simulation technology. In this article, a data-driven approach is proposed to analyzing the trajectory of urban electricity CEs based on empirical mode decomposition, which suggests combining macro-energy thinking and big data thinking by removing the barriers among power systems and related technological, economic, and environmental domains. Based on multisource heterogeneous mass data acquisition, effective secondary data can be extracted through the integration of statistical analysis, causal analysis, and behavior analysis, which can help construct a simulation environment supporting the dynamic interaction among mathematical models, multi-agents, and human participants.