A Full Life Cycle Prediction Method for Supply Chain Carbon Emissions Based on Integrated Deep Learning

In order to refine the company's carbon peak target indicators, establish a dynamic monitoring system for the indicators, and track the progress of the company's and each unit's emission reduction targets, this paper proposes a full life cycle prediction method for supply chain carbon emissions based on integrated deep learning. The method adopts a combination of gray prediction model and deep learning technology to model and predict the carbon emission process from the three-dimensional data of supply chain, delivery chain and design chain. In the test results, the upper limit of the relative error is no more than 8%, and the CNN-GM-LSTM model is more accurate in prediction compared with the GM (1, N) model and the CNN-LSTM hybrid model.