摘要
利用人工神经网络的非线性映射特性,以典型的煤样主要煤级指标做训练样本,对网络进行训练,对其中各煤化阶段的指标的权值和神经元内部的阈值沿误差下降的方向不断修正,最终达到了精度要求。该煤化阶段划分方法在实际工作中具有一定参考价值和指导意义。
Based on non - linear mapping of the artificial neural network, with the coal - level indicators of typical coal samples as the training samples, the paper analyses the coal rank dividing. The conclusion has some reference value and guiding significance in the practical work.
出处
《陕西煤炭》
2008年第6期65-67,共3页
Shaanxi Coal
关键词
人工神经网络
煤级指标
权值
误差
artificial neural network
coal rank index
weight
error
