Prediction of CO2 emission for the central European countries through five metaheuristic optimization techniques helping multilayer perceptron
| Autor: | Hossein Moayedi, Azfarizal Mukhtar, Serhan Alshammari, Mohamed Boujelbene, Isam Elbadawi, Quynh T Thi, Mojtaba Mirzaei |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2024 |
| Předmět: | |
| Zdroj: | Engineering Applications of Computational Fluid Mechanics, Vol 18, Iss 1 (2024) |
| Druh dokumentu: | article |
| ISSN: | 19942060 1997-003X 1994-2060 |
| DOI: | 10.1080/19942060.2024.2327437 |
| Popis: | One of the most significant issues in urban design is energy-related CO2 emissions, which are rising quickly as cities expand. The GDP of the Central European countries (from 1990 to 2016) based on several energy sources, such as coal, oil, natural gas, and renewable energy, are used as inputs in this study. To develop a reliable predictive network considering the problem complexity, multilayer perceptron (MLP) is combined with several nature-inspired optimization algorithms, namely, black hole algorithm (BHA), future search algorithm (FSA), backtracking search algorithm (BSA), biogeography-based optimization (BBO), and shuffled complex evolution (SCE). By applying the approaches mentioned above to the synthesis of the MLP, the recommended BBO, BHA, BSA, FSA, and SCE ensembles are obtained. A series of parametric studies are performed to improve the effectiveness of the employed models. It is found that, by combining the BBO, BHA, BSA, FSA, and SCE algorithms, the MLP's accuracy is increased. The result from this parametric analysis showed that SCE and BBO perform better than the other three algorithms as the CO2 emission was computed with the highest level of accuracy using R2 = 0.9999 and 0.9998, RMSE = 1.6781 and 2.0539 for SCE, and R2 = 0.9999 and 0.9998, RMSE = 1.8689 and 2.3833 for BBO. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|