基于Logistic改进模型的中国天然气产量预测

Prediction for Chinese Natural Gas Production:Based on a Developed Logistic Model

Logistic预测模型在能源预测领域有广泛的应用,但Logistic原模型的能源产量、储量等预测结果与实际值差异较大。其原因是在大多研究中将Logistic原模型中的“环境容量”和“种群内禀增长率”两个参数假设为常量,而实际上这两个参数是随时间变化的。本文通过将这两个常量参数函数化以改进模型。基于奇异摄动相关理论,使用匹配展开法设计了模型参数求解方法。以中国天然气历年的可采储量和产量数据进行实证研究,并与几种常用预测模型对比,验证改进模型的预测能力。结果表明,相较于Logistic原模型,改进后的模型明显地提高了预测精确度。最后运用改进的Logistic模型预测了我国2019-2036年的天然气产量,在2028年天然气产量达到峰值为3311.3亿立方米。

Logistic model is one of the most accepted prediction models, but its utilization in predicting natural gas production is often limited due to its intrinsically fixed environment threshold and population growth rate. In this paper, these two parameters are functionalized to be time-dependent according to the characteristics of natural gas production. Based on singular perturbation theory, the matching expansion method is applied to solve this time-dependent Logistic prediction model. China’s natural gas production is used as an empirical case, and comparisons between traditional prediction models are conducted. Results reveal that the developed Logistic model can effectively improve the prediction, and according to this model, China’s natural gas production will reach the peak of 331.13 billion cubic meters by 2028.