CarbonKG: industrial carbon emission knowledge graph-based modeling and application for carbon traceability of complex manufacturing process
Tao Wu,Jie Li,Jinsong Bao,Liu Qiang,Zhengkai Jin,Jun Gao
DOI: https://doi.org/10.1115/1.4065166
IF: 2.3
2024-03-26
Journal of Computing and Information Science in Engineering
Abstract:Abstract In the context of the 2030 carbon dioxide emissions peak target, the green transformation of the manufacturing industries is the only way to achieve the carbon neutrality target. However, the long supply chain and complex manufacturing process from raw material to the final manufacturing product makes it challenging to analyze the flow and distribution of the carbon footprint for the whole life cycle. The paper focuses on the manufacturing process, and proposes a modeling method for carbon emission quantification and prediction using industrial intelligent technologies such as process mining and knowledge graph. Firstly, it is crucial to establish comprehensive carbon emission quantification model for each step, including the energy flow, material flow, personnel flow, and carbon flow in the manufacturing process. Secondly, based on the quantification model, an industrial carbon emission knowledge graph-based model (CarbonKG) is proposed to capture and calculate the carbon emissions at each production process. Moreover, employing process mining technology enables analysis of the global carbon emission distribution and flow within the manufacturing process. Finally, this research presents a two-stage process carbon emissions prediction method based on the CarbonKG. In the first stage, a local order carbon graph (LCG) to be predicted is initialized, and a model is established to retrieve similar cases. Subsequently, a graph matching model is developed to identify the TOPK similar order cases, using their carbon emission values for comprehensive prediction. Furthermore, the effectiveness of the proposed method in the paper was verified using actual production process data from industrial companies.
engineering, manufacturing,computer science, interdisciplinary applications