基于小波去噪改进神经网络拱顶下沉预测研究

摘要

隧道拱顶下沉监测数据中含有大量的随机误差，为了消除或者消弱随机误差的干扰，本文对实测数据进行小波去噪，使数据更真实性。针对传统BP神经网络预测精度差、收敛慢的问题，通过改进的BP神经网络对去噪的数据进行预测。实验结果表明，并与传统BP神经网络相对比，小波去噪的改进神经网络收敛速度加快，精度提高，预测效果显著提高，适用于拱顶下沉的预测研究。%Vault sink of tunnel contains a lot of random error. In order to eliminate or weaken interference of random error, the measured data was processed by wavelet de-noising that made the data more authenticity in the paper. Aiming at problems such as poor precision and slow convergence about BP neural network prediction, de-noising data was predicted by the improved BP neural network, which compared with traditional BP neural network. Experimental results showed the improved neural network of wavelet de-noising made convergence rate accelerate, accuracy improve, prediction result significantly enhance, it was true to prediction research of vault sink.