论文部分内容阅读
隧道工程围岩的级别是隧道围岩稳定性的尺度,施工期间的隧道围岩分类的确定是最为基础、也是最为重要的内容。本文将粗糙集、小波神经网络和围岩分类有机结合起来,对白鹤隧道围岩分类进行识别研究。结果表明:用经过粗糙集约简后的样本集作为神经网络的训练样本集,有效地简化了神经网络的结构,减少了训练步数与训练时间,并提高了网络的学习速度和判断准确率;经过粗糙集约简后的WNN判别准确率最高,识别结果更接近专家质量评价法;而BP网络判别结果与专家质量评价法相差较大;总体上,小波神经网络预判的结果要比BP神经网络预判的结果精度要高,约简后要比约简前的精度要高。
The grade of tunnel surrounding rock is the measure of the stability of tunnel surrounding rock. The classification of tunnel surrounding rock during construction is the most basic and the most important content. This paper combines rough set, wavelet neural network and surrounding rock classification to identify the classification of surrounding rock of Baihe Tunnel. The results show that using the sample set reduced by rough set as the training sample set of neural network can effectively simplify the structure of neural network, reduce training steps and training time, and improve the learning speed and accuracy of the network. After rough set reduction, the accuracy of WNN identification is the highest, and the recognition result is closer to the expert quality assessment method. However, BP network discrimination results are greatly different from expert quality evaluation methods. Generally speaking, the results of WNN prediction are better than those of BP neural network The accuracy of the result of the pre-judgment is higher, and the reduction is more accurate than that of the pre-reduction.