吹填淤泥浅表层快速加固处理技术及神经网络沉降预测方法

Shallow Surface Rapid Reinforcement Technology of Dredger Fill and Artificial Neural Network Settlement Prediction Method

选取厦门港海沧港区软基处理工程为例,针对试验区吹填淤泥的工程特性,运用浅表层快速加固处理技术对其进行加固处理,详细介绍了该项技术的技术思路、作用机理以及与真空预压法的区别,评价这项技术的加固效果。然后利用Matlab软件编制了BP网络、RBF网络,Elman网络时序预测模型程序对加固后试验区的沉降量进行预测研究,从预测结果看,Elman网络的预测精度最高,其次为BP网络,相对最差的是RBF网络,Elman网络的反馈型网络结构使其具有更好的时序预测能力。

The author took the soft foundation＇treatment project of Haicang Port Area of Xiamen Port for an instance, and adopted reinforcement treatment by shallow surface rapid reinforcement technology based on the engineering characteris- tics of dredger fill in experiment plot. In the paper, it also introduced the technical ideas, mechanism of the technology and the difference with vacuum preloading method in detail, in order to evaluate the reinforcement effect, in the mean- time, BP neural network, RBF neural network and Elman neural network timing prediction model program have been es- tablished by using MATLAB to predict and research the settlement after reinforcement in this plot. From the prediction re- sults, the highest prediction accuracy is Elman neural network, the second is BP neural network, the relative worst is the RBF network. The feedback network structure of Elman makes it have better timing prediction capability.