Scholarlink

<p>A novel grey model, termed the variable adjacent neighbor weight grey model, VANWGM(1,1), is optimized using a genetic algorithm with a bi-objective approach to balance the trade-off between training (seen) and testing (unseen) data, thereby improving forecasting accuracy. The model is validated using two real-world datasets with different data patterns, demonstrating its performance. A rolling mechanism is also incorporated to handle relatively longer time series. Furthermore, the proposed model is compared with various existing grey models, confirming its competitive predictive performance and its ability to produce highly accurate results. Historical data on China’s fuel ethanol production, consumption, and blended fuel ethanol with gasoline from 2015 to 2023 are used to develop and evaluate the proposed VANWGM(1,1) model. Based on the forecast results, China’s fuel ethanol production is projected to reach 7,007 million liters by 2030. However, to meet the nationwide E10 (a blend of 10% fuel ethanol and 90% gasoline) biofuel mandate, an additional 19,979 million liters of fuel ethanol will be required. This gap can be addressed through increased domestic production, imports, or a combination of both. Finally, some possible recommendations are proposed to help decision-makers in China achieve a sustainable E10 biofuel mandate.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Variable Adjacent Neighbor Weight Grey Model Optimized by Genetic Algorithm and Its Application: Towards Achieving China’s E10 Mandate

  • 期刊论文

  • 发表日期：

    2025年5月9日

查看全文
  • Essam Kaoud,
  • Emad Alenany,
  • Islam M. Sharafeldin

摘要

A novel grey model, termed the variable adjacent neighbor weight grey model, VANWGM(1,1), is optimized using a genetic algorithm with a bi-objective approach to balance the trade-off between training (seen) and testing (unseen) data, thereby improving forecasting accuracy. The model is validated using two real-world datasets with different data patterns, demonstrating its performance. A rolling mechanism is also incorporated to handle relatively longer time series. Furthermore, the proposed model is compared with various existing grey models, confirming its competitive predictive performance and its ability to produce highly accurate results. Historical data on China’s fuel ethanol production, consumption, and blended fuel ethanol with gasoline from 2015 to 2023 are used to develop and evaluate the proposed VANWGM(1,1) model. Based on the forecast results, China’s fuel ethanol production is projected to reach 7,007 million liters by 2030. However, to meet the nationwide E10 (a blend of 10% fuel ethanol and 90% gasoline) biofuel mandate, an additional 19,979 million liters of fuel ethanol will be required. This gap can be addressed through increased domestic production, imports, or a combination of both. Finally, some possible recommendations are proposed to help decision-makers in China achieve a sustainable E10 biofuel mandate.