Achieving the carbon intensity target of China: A least squares support vector machine with mixture kernel function approach
By Bangzhu Zhu, Shunxin Ye, M. Jiang, Ping Wang, Zhanchi Wu, Rui Xie, Julien Chevallier, Yi-Ming Wei. Applied Energy, 2019
113 citations recorded by the source corpus.
This study proposes a novel least squares support vector machine with mixture kernel function-based integrated model for achieving the China’s carbon intensity target by 2020 from the perspective of industrial and energy structure adjustments. Firstly, we predict the industrial and energy structures by the Markov Chain model and scenario analysis, GDP by scenario analysis,and energy consumption by introducing a novel least squares support vector machine with mixture kernel function in which particle swarm optimization is employed for searching the optimal model parameters. Secondly, we deduce the carbon intensities and contribution potentials of industrial and energy structure adjustments to achieving the carbon intensity target by 2020 under 27 combined scenarios. Under 27 combined scenarios, carbon intensity in 2020 will decrease by 48.37%–52.62% compared with that of 2005. The scenario with GDP low-speed growth, industrial structure medium adjustment and energy structure major adjustment, will be the preferred path to achieving the carbon intensity target, in which the carbon intensity in 2020 will be 6.62 t/104 Yuan, declined by 51.73% compared with that of 2005. The obtained results also show that, compared with the least squares support vector with single radial basis and polynomial kernel functions, and cointegration equation models, the proposed least squares support vector with mixture kernel function can achieve a higher forecasting accuracy for energy consumption. The contribution potential of industrial structure adjustment is greater than that of energy structure adjustment to achieving the carbon intensity target.
Subfields: Economics and Econometrics, Environmental Engineering, Management Science and Operations Research
View Paper: https://www.semanticscholar.org/paper/a7950d7025c2cf9af8b70b493356d05c7ab72232
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Economics and Econometrics
In Environmental Engineering
Citations in the corpus
- Short-term CO2 emissions forecasting based on decomposition approaches and its impact on electricity market scheduling (2020, methodology): “[21] utilized least squares support vector machine (LSSVM) with the mixture kernel function based prediction model to set a CO2 intensity target by the year 2020 based on industrial and energy structures.”