中文摘要：

为了获得综放开采现场用以分类煤岩的有效的特征向量和分类模型,通过已有的设备及设计采集方案,对综放开采现场的煤岩声压信号进行了采集;并对获取的声压信号进行时域分析,得到时域特征向量并作为神经网络的输入向量;利用主元分析（简称PCA）,减少时域特征间的相关性,降低神经网络输入向量的维数;然后设计BP神经网络模型,通过比较梯度下降法与Levenberg-Marquard算法,得知基于LM训练法耗时明显小于梯度下降法。最后对比进行PCA与未进行PCA的LM算法的BP神经网络煤岩识别结果,得到PCA与LM算法的BP神经网络结合的方式识别准确率高且耗时短。

英文摘要：

To acquire the effective feature vectors and classification model for the coal- rock identification for fully mechanized top coal caving face,the acoustic pressure signals of coal and rock are collected from the site. By analyzing the acquired acoustic pressure signals in time domain,the feature vectors are acquired and used as the input vectors of neural network. Principal component analysis（ PCA） method is used to reduce the correlation between time- domain characteristics and the dimensions of the input feature vectors of neural network. Then the BP neural network model is designed. The comparison between the gradient descent method and Levenberg- Marquard method shows, the later consumes much less time. The neural network based on the LM method combing with PCA obtains higher identification accuracy and consumes less time than that without PCA.