基于ARMA-ARCH模型的自备电厂煤气供入量变化趋势预测研究  

Research on Forecasting Varying Tendency of Gas Supply in Self- Provided Power Plant Based on ARMA- ARCH

作　　者：孟华[1] 王建军[1] 王华[1] 范国锋[1] 李红娟[1] 

机构地区：[1]昆明理工大学冶金节能减排教育部工程研究中心,云南昆明650093

出　　处：《昆明理工大学学报（自然科学版）》2014年第3期66-72,共7页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：国家自然科学基金项目(51066002/E060701);NSFC-云南联合基金项目(U0937604)

摘　　要：为了掌握钢铁企业自备电厂煤气供入量的变化趋势,基于采样数据建立了自回归移动平均(ARMA)模型,利用拉格朗日乘数法(LM)检验出ARMA模型残差存在自回归条件异方差(ARCH)效应,建立ARMA-ARCH模型.分别使用ARMA模型和ARMA-ARCH模型进行短期预测,并比较两者的精度.最后基于概率分布对扰动项进行统计分析,得到生产中稳定性差是导致扰动项大的主要原因,与实际生产相吻合.研究表明,ARMA-ARCH模型的预测精度较高,预测误差为4.11%,能够较为准确地预测出钢铁企业自备电厂煤气供入量变化趋势,对实际生产中频繁调节锅炉开关和优化调度决策有着重要的作用.Varying tendency forecasting is very important to the residual gas supply of self-provided power plant in iron and steel industry.By applying the Eviews software,ARMA （Auto-Regressive Moving Average）model of varying tendency for residual gas supply is built.Through ARCH （Auto-Regressive Conditional Heteroskedastic） effect tests of the residual of ARMA model by Lagrange multiplier,the corresponding ARMA-ARCH model is also set up.The varying tendency series are forecasted by using ARMA model and ARMA-ARCH model respectively. Forecasting precision of ARMA model and ARMA -ARCH model is then compared.In this model,probability distribution is used to analyze the residual series in combination with the real production.The results indicate that the proposed model has a well -pleasing forecast performance with a Mean Abs.Percent Error （MAPE ）of 4. 1 1%.The case study shows that the model performs well to forecast the varying tendency,which can be used to keep production balance and ensure the pipeline network pressure in a safe range.

关 键 词：趋势预测 ARMA模型 ARCH效应 概率分布 

分 类 号：TF543.2[冶金工程—钢铁冶金]