摘要
针对传统的BP神经网络模型在数据拟合方面存在网络收敛速度慢,预测精度不高的缺点,提出一种改进的BP神经网络模型的方法。利用改进后的模型对某一地铁隧道变形监测数据进行分析和预报,并结合MATLAB软件编写的数据处理程序实现改进前后两种模型对同一数据处理结果的对比分析,验证改进后模型的有效性和可靠性。
In order to overcome the low forecast accuracy and slow convergence of traditional BP neural network model in data fitting,proposed an improved BP neural network method. Using improved model to analysis and forecast a subway tunnel deformation monitoring data,Combined with MATLAB data handler program to achieve comparison of the two models of thesame before and after the data processing results,Verify the effectiveness and reliability of the improved model.
出处
《城市勘测》
2017年第1期137-141,共5页
Urban Geotechnical Investigation & Surveying
基金
江西省教育厅教改立项课题(JXJG-15-52-10)