Carbon emission prediction method and device based on multidimensional reference quantity and time series

Abstract

The present invention provides a carbon emission prediction method and device based on multi-dimensional reference quantities and time series, including: calculating based on carbon emission data at each previous time point within a preset time period to obtain a first basic carbon emission prediction value; Set the economic data change trend and the industrial data change trend of the previous time point in the time period to obtain the second economic weight value and the second industrial weight value; for the second economic weight value, the second industrial weight value and the first basic carbon Emission forecast value is calculated to obtain the second basic carbon emission forecast value; based on the calculation of the second economic weight value, the second industrial weight value and the characteristic data of the forecast site, the offset interval of the forecast site is obtained; For multiple future time points, the carbon emission data corresponding to multiple future time points are predicted based on the second basic carbon emission prediction value and the offset interval, and the carbon emission interval data corresponding to each future time point is obtained.

Description

Carbon emission prediction method and device based on multi-dimensional reference quantities and time series technical field The present invention relates to carbon emission technical field, relate in particular to a kind of carbon emission based on multidimensional reference quantity and time series Prediction method and apparatus. Background technique [0002] Based on the existing literature, the prediction of carbon emissions is based on the impact of GDP, industrial added value and energy activities on carbon emissions. to analyze and predict carbon emissions, such as the commonly used classical models are Kaya identity, MERGE model and IPAC models, etc.; there are also models built by data analysis methods to predict carbon emissions. Common models include multiple Neural network, DDEPM model and GM (1, 1) model, etc., these models or methods can all improve the measurement and prediction of carbon emissions. For effective theoretical analysis and feasible veri...

AI Technical Summary

Problems solved by technology

[0004] (1) Existing forecasting models mostly rely on economic and industrial data to assess and predict carbon emissions. There will be problems such as insufficient reference data and low prediction accuracy. The reason for this problem is the prediction model. The reference dimension of carbon emissions is less

[0005] (2) Nowadays, most of the data forecasts for carbon emissions are for the whole country or even the whole world, and the forecasts for regional carbon emissions are less, and for the forecast of regional carbon emissions, the relevant energy structure of the region should be considered And energy consumption policy analysis is less, so the forecast accuracy of provincial carbon emissions is insufficient, and the forecast of carbon emissions for regions is less

[0006] (3) Moreover, when predicting carbon emissions, they only predict one future time point, and cannot predict multiple consecutive time points, resulting in poor periodicity and many limitations in carbon emission prediction

[0007] Based on the above problems, there is an urgent need for a carbon emission prediction method with multiple reference dimensions and a large prediction time span

Images