共 115 条
Renewable energy strategy analysis in relation to environmental pollution for BRICS, G7, and EU countries by using a machine learning framework and panel data analysis
被引:3
作者:
Cristea, Dragos Sebastian
[1
]
Zamfir, Cristina Gabriela
[1
]
Simionov, Ira Adeline
[2
,3
]
Fortea, Costinela
[1
]
Ionescu, Romeo Victor
[4
]
Zlati, Monica Laura
[1
]
Antohi, Valentin Marian
[1
,5
]
Munteanu, Dan
[6
]
Petrea, S. M.
[1
,2
]
机构:
[1] Dunarea de Jos Univ Galati, Fac Econ & Business Adm, Dept Business Adm, Galati, Romania
[2] Dunarea de Jos Univ Galati, Fac Food Sci & Engn, Dept Food Sci Food Engn Biotechnol & Aquaculture, Galati, Romania
[3] Dunarea de Jos Univ Galati, Fac Automat Control & Syst Engn, Dept Automat Control & Elect Engn, Galati, Romania
[4] Univ Galatzi, Fac Jurid Social & Polit Sci, Dept Adm Sci & Reg Studies, Galati, Romania
[5] Transylvania Univ, Fac Econ Sci & Business Adm, Dept Finance Accounting & Econ Theory, Brasov, Romania
[6] Dunarea de Jos Univ Galati, Fac Automatic Control & Syst Engn, Dept Comp & Informat Technol, Galati, Romania
基金:
芬兰科学院;
关键词:
renewable energy;
environmental pollution;
sustainable development;
BRICS;
G7;
EU;
ECONOMIC-GROWTH EVIDENCE;
NATURAL-GAS CONSUMPTION;
CO2;
EMISSIONS;
NONRENEWABLE ENERGY;
EMPIRICAL-EVIDENCE;
OIL CONSUMPTION;
NEXUS;
QUALITY;
D O I:
10.3389/fenvs.2022.1005806
中图分类号:
X [环境科学、安全科学];
学科分类号:
08 ;
0830 ;
摘要:
The present research uses machine learning, panel data and time series prediction and forecasting techniques to establish a framework between a series of renewable energy and environmental pollution parameters, considering data for BRICS, G7, and EU countries, which can serve as a tool for optimizing the policy strategy in the sustainable energy production sector. The results indicates that XGBoost model for predicting the renewable energy production capacity reveals the highest feature importance among independent variables is associated with the gas consumption parameter in the case of G7, oil consumption for EU block and GHG emissions for BRICS, respectively. Furthermore, the generalized additive model (GAM) predictions for the EU block reveal the scenario of relatively constant renewable energy capacity if gas consumption increases, while oil consumption increases determine an increase in renewable energy capacity until a kick point, followed by a decrease. The GAM models for G7 revealed the scenario of an upward trend of renewable energy production capacity, as gas consumption increases and renewable energy production capacity decreases while oil consumption increases. In the case of the BRICS geopolitical block, the prediction scenario reveals that, in time, an increase in gas consumption generates an increase in renewable energy production capacity. The PCA emphasizes that renewable energy production capacity and GHG, respectively CO2 emissions, are highly correlated and are integrated into the first component, which explains more than 60% of the variance. The resulting models represent a good prediction capacity and reveal specific peculiarities for each analyzed geopolitical block. The prediction models conclude that the EU economic growth scenario is based on fossil fuel energy sources during the first development stage, followed by a shift to renewable energy sources once it reaches a kick point, during the second development stage. The decrease in renewable energy production capacity when oil consumption increases indicates that fossil fuels are in trend within the G7 economy. In the case of BRICS, it is assumed that gas consumption appears because of increasing the industrial capacity, followed by the increase of economic sustainability, respectively. In addition, the generalized additive models emphasize evolution scenarios with different peculiarities, specific for each analyzed geopolitical block.
引用
收藏
页数:29
相关论文