基于神经网络的放煤巷道支护方案优选

OPTIMAL SELECTION OF SUPPORT PATTERNS FOR CAVING ROADWAY BASED ON NEURAL NETWORK

根据急倾斜煤层巷道放顶煤采煤法的特点,分析了影响放煤巷道稳定性的因素,建立了基于改进型BP神经网络的放煤巷道支护方案优选和巷道稳定性预测模型,为放煤巷道的支护设计和急倾斜煤层放顶煤开采的高产高效提供科学依据和可靠保障。应用结果表明:采用改进型BP神经网络建立的模型,收敛速度快,预测精度高,有较大的实用价值。

According to the features of roadway sub-level caving in steep seam,the factors influencing the stability of caving roadway are analyzed,and the model for selecting optimal support patterns and predicting stability of caving roadway is established based on improved BP neural metwork. The application of the model can provide scientific basis for support design of caving roadway and reliable guarantee for high product and high efficiency of roadway sub-level caving in steep seam. Application results show that the presented model with improved BP neural metwork is of high learning speed,good prediction precision and significant practical value.