文档介绍:第25卷第9期 2008年9月公路交通科技 Journal Highway and Transportation Research and Development 文章编号:1002-0268(2008)09-0346-04 灰色预测模型(GM(1,1))预测精度浅析;E4,-1,沈鹃2 (,重庆400067;,四川绵阳621000) 摘要:通过实例,对GM(1,1)预测模型的精度与原始数据量的关系、与预测步的关系、与发展系数的关系、与原始数据的规律性的关系等方面进行了分析。通过分析发现:GM(1,1)模型的预测精度较高,与原始数据量之间没有明显的正相关或负相关关系;随着预测步的增加,预测精度明显降低;发展系数一a满足o<一a≤,发展系数为 ;原始数据的代表性和规律性对预测结果的精度影响较大。关键词:灰色系统,GM(1,1)模型,预测模型,预测精度中图分类号: 文献标志码:A Analysis uracy ofGrey Forecast Model(GM(1,1)) KONG Ling-yunl,SHEN Juan2 ( Research&Design Institute,Chongqing 400067,China; University ofScience and Technology,Mianyang 621000,China) Abstract Through example,the relationship between uracy theGM(I,I)prediction model and thenumber the original data,the relationship between the accuracy and Prediction Step,the relationship between the accuracy and development coefficient,the relationship between uracy and the regularity of theoriginal dataand SO on have been following Call befound:the accuracy theGM(1,1)prediction model no positive ornegative correlationbetween uracy themodel and thenumber theoriginal theprediction stepincreasing, uracy decreases thedevelopment coefficient—ameet‰≤ accuracy highest thepoint representation and regularity theoriginal datahavegreatimpact on