Power Carbon Emission Right Allocation Optimization Model Based on Power Data Mining and BAS-LSTM

Abstract

Aiming at the problems of uneven distribution and uncoordinated distribution amount, the optimization model of power carbon (C) footprint rights distribution based on power data mining and BAS-LSTM. Collect the power generation data and preprocess the data, extract the power C footprints data through K-means clustering algorithm, use it as the input data to build the power C footprints prediction model based on LSTM (Long term and Short Term Memory Network), optimize the number of hidden layers and threshold of LSTM through BAS algorithm, and use the LSTM C footprints prediction model with the best parameters; According to the principle right distribution, the thermal power units with the lowest economic cost of comprehensive power generation of each power plant unit and no distinction between the size units are taken right distribution targets of power generation units and power plant groups, and an optimization model right distribution of power is established right distribution scheme. The experiment shows that the obtained power C footprint right allocation scheme can enable most power plants to obtain satisfactory C footprint quotas. The resource allocation is very reasonable, and is more harmonious and scientific than the traditional allocation scheme.