Title: Hierarchical classification of dynamic carbon emission factors based on improved support vector machine

Authors: Chenghao Xu; Baichong Pan; Weixian Che

Addresses: Research Center of Grid Planning, Guangdong Power Grid Co., Ltd., Guangzhou, Guangdong, 510080, China ' Research Center of Grid Planning, Guangdong Power Grid Co., Ltd., Guangzhou, Guangdong, 510080, China ' Research Center of Grid Planning, Guangdong Power Grid Co., Ltd., Guangzhou, Guangdong, 510080, China

Abstract: In order to solve the problems of low factor coverage and low factor comprehensiveness existing in the traditional hierarchical classification method of carbon emission factors, a hierarchical classification method of dynamic carbon emission factors based on improved support vector machine is proposed. Firstly, collect dynamic carbon emission data and pre-process the data, calculate the contribution of each factor to the change of total carbon emission according to LMDI decomposition method, and determine the weight of each dynamic carbon emission factor. Introduce kernel function into OC-SVM algorithm in improved support vector machine, map dynamic carbon emission factors to high-dimensional space, and update the optimal hyperplane position with disturbance factors to realise hierarchical classification of dynamic carbon emission factors. The experimental results show that the factor coverage rate of the proposed method is above 90%, the highest factor comprehensiveness can reach 95%, and the practical application effect is good.

Keywords: improving support vector machine; dynamic carbon emission factors; hierarchical classification; OC-SVM algorithm; LMDI decomposition method; Kernel function.

DOI: 10.1504/IJETP.2025.144307

International Journal of Energy Technology and Policy, 2025 Vol.20 No.1/2, pp.163 - 181

Received: 29 Apr 2024
Accepted: 08 Jul 2024
Published online: 05 Feb 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article