论文部分内容阅读
由于地下金属矿床地质与开采条件的复杂性,影响岩层移动的因素错综复杂且相互影响,使得对岩层移动的预测具有很大的不确定性。大量的样本数据减慢了神经网络的训练速度,并且使得神经网络不稳定。将主成分分析(PCA)与Elman网络相结合构建模型,对地下矿山岩层移动角进行预测研究。利用主成分分析对原始数据进行预处理,提取原信息的主成分,将输入变量减少且互不相关,提高神经网络训练速度;用Elman网络对训练样本进行训练,进而利用训练好的网络对预测样本进行预测,与不采用PCA时的预测结果相比,采用PCA的预测结果更为准确,通过期望输出与实际输出的对比,相对误差都在5%以内,其预测的结果精度高,表明了PCA与Elman网络相结合对地下矿山岩层移动进行研究是可行的。
Due to the complexity of geology and mining conditions of underground metal deposits, the factors that affect the movement of rock are complex and influence each other, making the prediction of rock movement with great uncertainty. A large amount of sample data slowed down the training speed of neural network and made the neural network unstable. Principal component analysis (PCA) and Elman network are combined to build a model to predict the movement angle of underground mine strata. The principal component analysis (PCA) is used to preprocess the original data, extract the principal components of the original information, reduce and irrelevant the input variables, and improve the training speed of the neural network; train the training samples with the Elman network, and then use the trained network to predict Compared with the predicted results without PCA, the predicted results using PCA are more accurate, and the relative errors are less than 5% by comparing the expected output with the actual output, and the accuracy of the prediction results is high The combination of PCA and Elman network is feasible to study the movement of underground mine strata.