Simulation Analysis of Energy Consumption and Carbon Emission Correlation Based on Unsupervised Feature Generation Algorithm

In order to carry out the carbon emission work more accurately, it is extremely necessary to clarify the relationship between energy consumption and carbon emission, therefore, we propose a simulation and analysis study on the correlation between energy consumption and carbon emission based on unsupervised feature generation algorithm. Carbon emission, carbon emission intensity and carbon emission density are taken as carbon emission features, and the energy volume for local consumption, processing and conversion inputs and outputs, losses, terminal consumption, balance difference and total consumption are analyzed from the perspective of energy flow, and the unsupervised matching of the mapping between carbon emission features and energy consumption is carried out by means of expansion convolution. In the simulation test scenarios, the design method provides the best fit between the carbon emission outputs of the constructed scenarios and the actual situation, and can effectively analyze the relationship between energy consumption and carbon emission.