摘要
地基沉降的机理十分复杂,难以用一种预测模型精确预测。结合某地基沉降的实际数据,采用时间序列分析法和BP神经网络法相结合的组合模型进行预测,并用马尔科夫理论对预测结果进行改进,得到了更可靠的结果。
Foundation settlement has very complex mechanism,a single method can hardly have high accuracy of the prediction.In this paper,we presented a combination model of ARIMA-BP neural network and did a prediction for a foundation settlement,then improve the predicted result by using Marcov theory and obtained the better result.
出处
《测绘地理信息》
2016年第4期32-36,共5页
Journal of Geomatics
基金
国家自然科学基金资助项目(41074025)~~
