Prediction of soft soil foundation settlement in Guangxi granite area based on fuzzy neural network model

2017 3rd International Conference on Environmental Science and Material Application

生态环境科学;材料科学

Luo, Junhui^1 ; Wu, Chao^2 ; Liu, Xianlin^1 ; Mi, Decai^1 ; Zeng, Fuquan^3 ; Zeng, Yongjun^3

Guangxi Communications Planning Surveying and Designing Institute Co., Ltd, Nanning

530029, China^1

Hunan Institute of Technology, Hengyang

421002, China^2

Guangxi Vocational and Technical College of Communications, Nanning, Guangxi, China^3

关键词: Approximation methods;    Calculation programs;    Compatibility methods;    Engineering designer;    Fuzzy neural network model;    Prediction and analysis;    Soft soil foundation;    T-S fuzzy neural network;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/108/3/032034/pdf
DOI  :  10.1088/1755-1315/108/3/032034

来源: IOP

【 摘 要 】

At present, the prediction of soft foundation settlement mostly use the exponential curve and hyperbola deferred approximation method, and the correlation between the results is poor. However, the application of neural network in this area has some limitations, and none of the models used in the existing cases adopted the TS fuzzy neural network of which calculation combines the characteristics of fuzzy system and neural network to realize the mutual compatibility methods. At the same time, the developed and optimized calculation program is convenient for engineering designers. Taking the prediction and analysis of soft foundation settlement of gully soft soil in granite area of Guangxi Guihe road as an example, the fuzzy neural network model is established and verified to explore the applicability. The TS fuzzy neural network is used to construct the prediction model of settlement and deformation, and the corresponding time response function is established to calculate and analyze the settlement of soft foundation. The results show that the prediction of short-term settlement of the model is accurate and the final settlement prediction result has certain engineering reference value.