水利水运工程学报
水利水運工程學報
수이수운공정학보
HYDRO-SCIENCE AND ENGINEERING
2010年
1期
90-94
,共5页
改进变维分形%边坡监测%预测%小数据量
改進變維分形%邊坡鑑測%預測%小數據量
개진변유분형%변파감측%예측%소수거량
improved variable dimension fractal%monitoring of slope%forecast%insufficient data%fractal theory
针对以往预测模型在数据少和噪音干扰下出现预测精度降低的问题,基于分形理论,尝试建立改进的变维分形预测模型,并以小湾工程边坡位移监测数据为例,选取D1、D2曲线作为预测模型的分形参数曲线,计算各曲线的分段分形维数,对位移进行预测,并分别用灰色模型GM(1,1)和BP神经网络进行对比预测.结果证明,这种方法充分利用了分形理论自相似性的特点,抗噪性强,能较好地应用于小数据量监测数据的预测,并且精度较高,有着良好的应用前景.
針對以往預測模型在數據少和譟音榦擾下齣現預測精度降低的問題,基于分形理論,嘗試建立改進的變維分形預測模型,併以小灣工程邊坡位移鑑測數據為例,選取D1、D2麯線作為預測模型的分形參數麯線,計算各麯線的分段分形維數,對位移進行預測,併分彆用灰色模型GM(1,1)和BP神經網絡進行對比預測.結果證明,這種方法充分利用瞭分形理論自相似性的特點,抗譟性彊,能較好地應用于小數據量鑑測數據的預測,併且精度較高,有著良好的應用前景.
침대이왕예측모형재수거소화조음간우하출현예측정도강저적문제,기우분형이론,상시건립개진적변유분형예측모형,병이소만공정변파위이감측수거위례,선취D1、D2곡선작위예측모형적분형삼수곡선,계산각곡선적분단분형유수,대위이진행예측,병분별용회색모형GM(1,1)화BP신경망락진행대비예측.결과증명,저충방법충분이용료분형이론자상사성적특점,항조성강,능교호지응용우소수거량감측수거적예측,병차정도교고,유착량호적응용전경.
Aiming at the problem of low precision caused by insufficient data and noise interruption,the paper attempts to set up and improve the forecasting model of variable dimension based on fractal theory. Meanwhile,with the monitoring data of slope displacement from the Xiaowan Project,and Curve D1,D2 as the fractal parameter curves of the forecasting model，it tends to predict the displacement by calculating the sub-fractal dimension,and comparing the forecast result with the gray model GM (1，1) and BP neural network. In this way,it makes good use of the self-similarity of the fractal theory and thus brings about a vast range of prospect for application due to its high precision and noise immunity.
