摘要
一般灰色预测模型因所需样本数据少、计算简单而在众多领域得到广泛应用,但其要求数据序列符合光滑离散函数且以数列的第一个数据点为定解条件,这就影响了一般灰色模型的应用范围和预测精度。文章引入阶跃函数以求改变原始数列的光滑度,以原始数据序列中每一点为已知条件分别求解预测公式和预测误差,以预测误差最少的预测公式为最佳预测模型。修正后的灰色预测模型不仅提高了预测精度,而且对具有突变点或断点的数据序列也有很好的预测效果。将新模型应用于中石化河南油田的油气生产成本预测,取得了较好的应用效果。
General grey forecasting models have been extensively applied in many areas because it is easily calculated and it deals with systems characterized by poor or lacking information. But the data of original data sequence must be taken at equal intervals and in consecutive order without bypassing any data. In addition, the first data of original data sequence must be served as the initial qualification. These affect its forecasting accuracy and application fields of general grey forecasting models. Aiming to extend the application of GM (1, 1) and improve the forecasting accuracy, two methods are referenced to create a new improved grey forecasting model in this paper. Stair function is introduced to enhance the smooth degree of the data sequence. And forecasting formula and forecasting residual are founded based on every point to find the best forecasting formula whose forecasting residual is least. Applying this improved model to forecast the oil and gas production cost of the Henan Oilfield, the result is much more accurate than that by traditional grey forecasting method. Thus, the improved grey forecasting model is proved to be more practicable.
出处
《天然气工业》
EI
CAS
CSCD
北大核心
2006年第8期149-150,153,共3页
Natural Gas Industry
关键词
石油
天然气
生产成本
预测
数学模型
精度
应用
oil, natural gas, production cost, forecast, mathematical model, accuracy