基于BP神经网络的底板破坏深度预测

Forecast of destroyed floor depth based on BP neural networks

在总结采场底板破坏深度预测方法和理论的基础上,结合大量实际资料分析,归纳出开采深度、煤层倾角、开采厚度、工作面长度、底板抗破坏能力和有无切穿型断层或破碎带6个方面是影响底板破坏深度的主要因素.根据全国典型突水案例,构建基于BP神经网络的底板破坏深度的预测模型,确定建立BP神经网络所需的输入样本和检验样本,运用Matlab软件对网络进行训练,得出了优化的网络模型,并根据建立的网络模型预测肥城煤田曹庄井田8812和9604工作面的底板破坏深度.通过与实测结果对比,证明该网络模型的计算结果比相关规程提供的底板破坏深度经验公式计算的结果更接近实际.

Based on summing up forecasting method and theory about destroyed floor depth of working face, combined with a lot of practical information analyses, the main six factors controlling destroyed floor depth were found out, which are mining depth, coal seam inclination angle, mining thickness, workfaee inclined length, coal floor anti-destructive capacity, fault or broken zone. According to the typical water-inrush cases in China coal field, forecast of destroyed floor depth based on BP neural networks was built and learning samples and test samples of neural network was determined. Optimistic network model was got by training with Matlab software. The destroyed floor depth of 8812 workface and 9604 workface in Caozhuang Mine, Feicheng coal field was forecast according to the established network model. By comparing the results of neural network model and the results of empirical formula provided by national regulations with the actual measurement results, the results obtained by neural network model are closer to reality than the results of empirical formula calculation.