In the Jining No.2 coalmine have well developed minor structures, thus greatly impacted coalmine production and safety. In allusion to difficult prediction of minor structure ahead of coal roadway advance, based on geological theories and mastered geological structure regular pattern, generalized coal seam dip angle, thickness, fissure type, gas accumulation, water inflow, temperature and bro-ken degree are main controlling factors of minor structures. Through measurement of structural surface and impacting factors in coal and rock variations in normal area, impacting area and destructed area, initially modeled minor structure prediction nonlinear network model based on BP artificial neural network, and using Matlab software carried out training and optimization to get optimized prediction model. Using the model carried out minor structure prediction on Nos.113L01, 113L02, and 113L06 working faces in No.11 winning district. The result has shown that the network model predicted results are tally with measured results.%济宁二号煤矿小构造发育，给煤矿生产及安全带来很大影响。针对煤巷掘进前方小构造预测难问题，根据地质理论和已掌握的地质构造规律，归纳出煤层的倾角、厚度、裂隙类型、瓦斯聚集量、涌水量、温度和破碎程度是影响小构造存在的主控因素，通过实测构造面和煤岩层中各影响因素在煤层正常区、影响区和破坏区的变化，初步建立了基于BP人工神经网络的小构造预测的非线性网络模型，并运用Matlab软件对模型进行训练和优化，得到了优化后的预测模型。并应用模型对十一采区的113下01、113下02和113下06工作面的小构造进行预测，结果表明该网络模型的预测结论与实测结果相符。
展开▼
