Carbon emission prediction method, device, equipment and storage medium

By screening historical carbon emission impact factors and building and adjusting carbon emission forecasting networks, the problem of inaccurate carbon emission forecasting in the existing technology has been solved, and an accurate prediction of the total future carbon emissions has been achieved.

CN114781272BActive Publication Date: 2025-03-18PINGAN INT SMART CITY TECH CO LTD
2 Cites 0 Cited by

Patent Information

Application Number
CN202210510895.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-05-11
Publication Date
2025-03-18
Estimated Expiration
2042-05-11

AI Technical Summary

Technical Problem

The existing carbon emission forecasting scheme cannot accurately extract the laws of carbon emission growth, resulting in the inability to accurately predict the total carbon emissions in the future.

Method used

A carbon emission prediction method is proposed. By screening the target impact factor from the preset historical carbon emission impact factor, a carbon emission prediction network is obtained, and based on the network, the input data and real-time carbon emissions are predicted, the network parameters are adjusted to form a prediction model, and then the total carbon emissions are predicted at the future moment.

Benefits of technology

Accurate prediction of the total carbon emissions at the future moment is achieved, the accuracy and efficiency of the prediction are improved, and the stability of the prediction model is ensured.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present invention relates to artificial intelligence, and provides a carbon emissions prediction method, device, equipment and storage medium. The method selects a target influencing factor from preset historical carbon emissions influencing factors, obtains multiple actual emissions at the first moment and input data corresponding to the target influencing factor, and obtains multiple real-time carbon emissions at the second moment, predicts the input data and real-time carbon emissions based on a carbon emissions prediction network, obtains target predicted emissions, adjusts the carbon emissions prediction network according to the actual emissions and the target predicted emissions, obtains a carbon emissions prediction model, predicts the influencing data and emissions based on the carbon emissions prediction model, and obtains the total carbon emissions at the predicted moment. In addition, the present invention also relates to blockchain technology, and the total carbon emissions can be stored in a blockchain.
Need to check novelty before this filing date? Find Prior Art