Wang Yong, Chi Pei, Nie Rui, Ma Xin, Wu Wenqing, Guo Binghong
School of Sciences, Southwest Petroleum University, Chengdu, 610500 Sichuan China.
School of Mathematics and Physics, Southwest University of Science and Technology, Mianyang, 621010 Sichuan China.
Soft comput. 2023;27(14):9321-9345. doi: 10.1007/s00500-023-08203-y. Epub 2023 Apr 21.
With the continuous depletion of global fossil energy, optimizing the energy structure has become the focus of attention of all countries. With the support of policy and finance, renewable energy occupies an important position in the energy structure of the USA. Being able to predict the trend of renewable energy consumption in advance plays a vital role in economic development and policymaking. Aiming at the small and changeable annual data of renewable energy consumption in the USA, a fractional delay discrete model of variable weight buffer operator based on grey wolf optimizer is proposed in this paper. Firstly, the variable weight buffer operator method is used to preprocess the data, and then, a new model is constructed by using the discrete modeling method and the concept of fractional delay term. The parameter estimation and time response formula of the new model are deduced, and it is proved that the new model combined with the variable weight buffer operator satisfies the new information priority principle of the final modeling data. The grey wolf optimizer is used to optimize the order of the new model and the weight of the variable weight buffer operator. Based on the renewable energy consumption data of solar energy, total biomass energy and wind energy in the field of renewable energy, the grey prediction model is established. The results show that the model has better prediction accuracy, adaptability and stability than the other five models mentioned in this paper. According to the forecast results, the consumption of solar and wind energy in the USA will increase incrementally in the coming years, while the consumption of biomass will decrease year by year.
随着全球化石能源的不断枯竭，优化能源结构已成为各国关注的焦点。在政策和资金的支持下，可再生能源在美国能源结构中占据重要地位。提前预测可再生能源消费趋势对经济发展和政策制定起着至关重要的作用。针对美国可再生能源消费年度数据少且变化大的特点，本文提出了一种基于灰狼优化算法的变权缓冲算子分数阶延迟离散模型。首先，运用变权缓冲算子方法对数据进行预处理，然后利用离散建模方法和分数阶延迟项概念构建新模型。推导了新模型的参数估计和时间响应公式，证明了结合变权缓冲算子的新模型满足最终建模数据的新信息优先原则。利用灰狼优化算法对新模型的阶数和变权缓冲算子的权重进行优化。基于可再生能源领域太阳能、总生物质能和风能的消费数据，建立了灰色预测模型。结果表明，该模型比本文提及的其他五个模型具有更好的预测精度、适应性和稳定性。根据预测结果，未来几年美国太阳能和风能的消费量将逐年递增，而生物质能的消费量将逐年下降。
