摘要
为提高利用季节性叠加趋势模型预测有较大波动性数据序列的预测精度,提出一种季节性叠加趋势—马尔柯夫组合预测新方法,并用于油菜平均产量的预测.采用浙江省诸暨市1949年到1996年的油菜平均每公顷产量数据建立一个季节性叠加趋势—马尔柯夫组合预测模型,对1997年到2003年的油菜平均每公顷产量进行预测,预测精度分别为:97·9%、97·9%、97·9%、97·9%、98·8%、97·7%和98·4%,远远高于季节性叠加趋势模型的预测精度:76·1%、68·9%、70·9%、97·9%、82·5%、76·9%和82·2%.该方法具有计算简单、精度高的特点.说明利用季节性叠加趋势—马尔柯夫组合预测模型可以大大提高具有周期趋势性和较大波动性数据序列的预测精度.
A new forecasting model combined the season tendency superposing and Markov forecasting methods together is presented to forecast the average yield of rapeseed. It has the merits of both simplicity of calculation and high forecasting precision to forecast data sequences with season tendency superposing and heavy random fluctuation. The forecasting model was based on historical data of the average yield of the rapeseed from 1949 to 1996 in Zhuji, Zhejiang, and forecast the average yield of the rapeseed from 1997 to 2003 in Zhuji, Zhejiang by the season tendency superposing-Markov forecasting model. The forecasting precision of season tendency superposing--Markov forecasting model from 1997 to 2003 was 97.9%, 97.9%, 97.9%, 97.9%, 98. 8%, 97.7%, 98.4% respectively, and in the season tendency superposing model, it was 76.1 %, 68.9 %, 70. 9%, 97.9%, 82.5%, 76.9%, 82.2% respectively. It shows that the season tendency superposing--Markov forecasting model can improve the forecasting precision highly when forecasting the data sequences wih season tendency superposing and heavy random fluctuation.
出处
《浙江大学学报(农业与生命科学版)》
CAS
CSCD
北大核心
2008年第3期309-314,共6页
Journal of Zhejiang University:Agriculture and Life Sciences
基金
国家自然科学基金资助项目(30671213)
教育部高等学校优秀青年教师教学科研奖励计划资助项目(02411)
关键词
季节性叠加趋势
马尔柯夫
预测
season tendency superposing
Markov model
forecast