【作者】 陈立伟；

【导师】 周英；

【作者基本信息】 河南理工大学 ， 采矿工程， 2007， 硕士

【摘要】 综放开采过程中如何提高综放工作面回采率是目前综放开采实践中存在的重大关键技术问题之一。针对这一难题,本项目以义煤集团千秋煤矿21181和21121工作面的地质条件和回采工艺为基础,采用人工神经网络预测和现场试验相结合的方法,寻求提高综放工作面回采率的途径。由于影响回收率的因素较多,且各因素之间的关系十分复杂,很难用确切的数学方程来描述。人工神经网络技术具有很强的非线性映射能力和自适应、自学习能力,特别适用于解决因果关系复杂的非确定性推理、判断、预测和分类等问题。故选用BP神经网络,对综放工作面回收率进行预测。在研究影响综放工作面回收率因素的基础上,本着全面、完整和可靠的原则,收集了潞安、兖州、徐州、铁法4个矿区的22个综放工作面回采率资料。进而运用BP神经网络,对综放工作面回收率进行了预测。经BP神经网络的预测和义马煤业集团千秋煤矿21181工作面顶煤回收率试验的相互印证,提出了提高工作面回采率的最优放煤工艺和最佳配套设备。同时,在21181和21121工作面进行了初、末采技术试验和综放工艺模式试验。最后,针对千秋矿的具体条件,从影响综放工作面回收率因素着手,在通过预测和现场试验的基础上,提出了提高综放工作面回采率的途径,包括减少初、末采和端头损失的技术途径,以及适合千秋煤矿综放工作面高产、高效、高回收率的最优放煤方式和最佳工艺模式。更多还原

【Abstract】 How to increase recovery rate of fully mechanized sub-level caving face in process of mining, which is among important and key technique problems what exist in practice of fully mechanized sub-level caving. Aim at this problem, this item seeks ways of increasing recovery rate of fully mechanized sub-level caving face, basing on geological conditions and mining craft of the 21181and 21121 working face, by associating Artificial Neural Network prediction with scene experiment. Owing to influence factors of recovery rate what are quiet excessive, and their collections are complex, which are hard to be depicted in way of math equations. Artificial Neural Network (ANN) technique has a very strong nonlinear ability and the self-adaptability. Especially, it is suitable for resolving the problems such as complex uncertain inference of the causalities, determination, forecast, classification and so on. So choosing BP Network to predicte recovery rate of fully mechanized sub-level caving face. Base on the research of effect factors of recovery rate of fully mechanized sub-level caving face, under principle of overall、integrity、reliant, the data of recovery rate of twenty-two fully mechanized sub-level caving face in Luan、Yanzhou、Xuzhou and Tiefa four diggings have been chose. Using BP neural network arithmetic, the relation model between recovery rate of fully mechanized sub-level caving face and effect factors has been founded, and their relation has been confirmed, moreover recovery rate of fully mechanized sub-level caving face have been forecasted. After prediction of the BP neural network and verifying by the experiments of top coal recovery rate in 21181 working face of the Qianqiu coal mine of Yima Coal Group, the caving craft and match equipments which increasing recovery rate of the fully mechanized sub-level caving face. At the same time, the experiments of first and end mining ways or craft models have been carried in the 21181 and 21121 working faces. Finally, Aim at the frongose situation of Qianqiu mine coal, from the aspect of ways on increasing recovery rate of fully mechanized sub-level caving face, base on prediction and scene experiment, it achieves approaches of increasing recovery rate of fully mechanized sub-level caving face, which include approaches of reducing first and end mining losing and end losing, as well as the most excellent caving mining ways and the best craft model that adapting to the high yield、high efficiency and high recovery rate of fully mechanized sub-level caving face of Qianqiu mine coal.更多还原