Method for predicting and verifying load electric quantity and carbon emission thereof

Abstract

The invention discloses a load electric quantity and carbon emission prediction and verification method, and belongs to the field of electric power economic dispatching. A load electric quantity and carbon emission prediction method comprises the following steps: counting high-carbon energy and power consumption of a user, and constructing an energy consumption data set; converting the consumption of the high-carbon energy into primary energy carbon emission, and constructing a primary energy carbon emission historical data set; taking the power consumption as a load data historical data set, converting the power consumption into power carbon emission, and constructing a power carbon emission historical data set; taking primary energy carbon emission and power load data as historical data, and establishing an LSTM prediction model and a GM prediction model of the primary energy carbon emission and the power load; carrying out dynamic weight integration, taking an integrated output result as a real prediction value, and outputting a prediction sequence; and converting the power load prediction value into a power carbon emission prediction value, completing reliability verification, and finally obtaining a total carbon emission prediction value.

Description

technical field [0001] The invention relates to the field of electric power economic dispatching, in particular to a method for predicting and verifying load electricity and its carbon emission. Background technique [0002] General industrial and commercial electricity prices continue to decline, and electricity has become an indispensable energy source for industrial production and a large amount of energy. In the absence of large-scale popularization of new energy power generation, large-scale use of electric energy will still indirectly generate greenhouse gas emissions. Therefore, when predicting carbon emissions, it is necessary to examine the predicted electricity consumption of industrial enterprises at the same time. In recent years, many scholars have conducted extensive research on the application of the LSTM method in user load forecasting, and relevant literature has also proved that it can achieve very good forecasting accuracy. At the same time, the GM metho...

AI Technical Summary

Problems solved by technology

Existing research and methods are mainly aimed at forecasting carbon emissions and total power load at the regional level. The granularity of forecasting modeling is not fine enough, and it can no longer meet the country's requirements for precise control of carbon emissions.

Images