深大基坑施工变形的智能控制技术

Intelligent control of deformation during excavation for large and deep foundation

利用神经网络 (ANN)和模糊控制 (FC)理论 ,采用预测控制的思想 ,建立了一套集基坑施工变形预测和控制于一体的智能化施工控制系统 ,该系统由神经网络预测器和模糊控制器组成。神经网络预测器对基坑变形进行连续滚动的多步预测 ,模糊控制器根据预测结果对施工参数进行决策控制。在MATLAB 5 .2平台支持下 ,研制了相应的基坑变形控制软件系统。实际应用结果表明 ,该智能控制系统对深基坑的安全施工过程具有较好的控制效果 ,真正做到了施工过程的实时、动态、智能化控制。软件系统操作界面简单、直观 ,便于实际工程应用。

This paper discusses application of artificial neural networks and fuzzy logical theory to control the deformation of diaphragm wall and ground settlements during excavation for large deep foundation. Based on MATLAB 5.2, an intelligent control system for safety of large foundation excavation was developed. The system has been tested on a large soft soil foundation excavation project in Shanghai. The results showed that the control system is very effective as a means of intelligent control for deformation of soil and diaphragm wall during foundation excavation.