Self-adaptive mining algorithm of power multisource carbon emission factors based on deep generation model

The strong duality of carbon emissions in the power industry requires feasible solutions for power carbon emission factors under any capacity configuration and uncertainty parameters. However, the trend of multi-source carbon emissions in the power industry is unstable, leading to relatively large mining errors and inconsistencies between the mining values and the true values. Therefore, this article proposes an adaptive mining algorithm for multi-source carbon emission factors in the power industry based on a deep generative model. Based on the significant correlation characteristics of carbon emission variables, a deep generative model is constructed, and radial basis functions are introduced as activation functions to output quantitative indicators of carbon emission factors. The standardized indicator data is applied to calculate the adaptive mining index of multi-source carbon emission factors in electricity, and the adaptive mining results of multi-source carbon emission factors in electricity are obtained. The experimental results show that the algorithm proposed in this paper has relatively small mining errors in different scenarios, and the mining values are consistent with the true value variation curve, which can more effectively mine the multi-source carbon emission factors of electricity.