灰色回归组合新模型在城市天然气需求预测中的应用

A new compound model of gray regression and its application to predicting urban natural-gas demand

季节、气温和人口变化等因素会影响城市天然气需求预测的准确性,而利用不同模型或方法对预测结果进行调整和修正将有助于提高预测准确率。为此,以灰色Verhulst模型和线性回归模型的拟合结果为基础,利用具有以新替旧功能的新陈代谢原理,构建灰色回归组合模型,以湖北省城市天然气需求量为例,进行需求量预测,以期为建立城市天然气供需平衡机制提供科学的理论数据支撑。研究结果表明:①组合预测模型能降低单一模型的预测风险,使预测结果更稳定可靠,对降低预测误差具有一定作用;②新陈代谢灰色回归组合预测模型对原预测模型具有一定的改进作用,能明显提高预测精度,预测结果可作为相关企业部门或政府部门做出科学决策的参考依据;③新陈代谢灰色回归组合预测模型拟合性好,预测精度高,预测结果更稳定可靠,可用于城市天然气中长期需求量预测,具有一定的实际应用价值。

It's to some extent difficult to accurately predict an urban natural-gas demand due to seasonal influence,and temperature and population change.Moreover,it's beneficial to improve this prediction accuracy by different models or methods to adjust and correct prediction results.Based on some results obtained from both gray Verhulst and linear regression models,another compound model of gray regression based on metabolic principle with metabolic functions was developed.By virtue of this model,the urban gas demand was predicted accurately so as to provide many scientifically the oretical data supporting for establishing gas supply-demand balance mechanisms.Then,Hubei Province was taken as an example to carry out gas demand prediction.Results show that(1)this compound model can reduce some prediction risks existing in single model and play a certain role in reducing prediction errors,and its results are more stable and reliable;(2)the metabolism model is modified from the original prediction model and can increase prediction accuracy significantly,and its results can be taken as the reference and basis for relevant enterprise departments and government departments to make decisions scientifically;and(3)the compound prediction model is characterized by good fitting,high accuracy,and stable and reliable results,and it can be used for short-term demand prediction of urban natural gas.Its application is practical.