CAJ | 학술논문

大坝早期变形是一个含有已知和未知因素的不确定的复杂过程,沉降变形计算和预测在水库运营管理中具有重要意义。针对大坝变形沉降曲线非平稳的特性,利用时间序列原理建模进行预测,并与实测数据进行比较。结果表明,时间序列预报模型较好地描述了变形监测点的变化规律,预测精度优于灰色模型,更适用于实际应用。
대패조기변형시일개함유이지화미지인소적불학정적복잡과정,침강변형계산화예측재수고운영관리중구유중요의의。침대대패변형침강곡선비평은적특성,이용시간서렬원리건모진행예측,병여실측수거진행비교。결과표명,시간서렬예보모형교호지묘술료변형감측점적변화규률,예측정도우우회색모형,경괄용우실제응용。
The dam′s early deformation is a complex process containing the known and unknown factors for uncertainty.The settlement deformation calculation and forecasting play an important role in reservoir operation and management.Here,according to the subsidence curves of non-stationary characteristics for dam deformation,the time series theory is used for modelling to predict the deformation of dam monitoring points and compared with the measured data.The final results show that the ARIMA model can better describe the change rule of monitoring points.The forecasting accuracy of the ARIMA is better than that of GM model,so the ARIMA model is fitter to be used in practice.