论文部分内容阅读
开采冲击危险煤层时,需要对能量大于等于104 J矿震发生的时间和地点进行预测。但目前所预测的结果不能满足现场要求。依据波兰一个长壁工作面的矿震事件,研究了神经网络系统中输入数据类型和形式对矿震预测效果的影响,确定了提高预测效果的方法。实践证明,采用神经网络技术可以对矿震危险性进行预测。
Mining the impact of dangerous coal seams, the need for energy greater than or equal to 104 J mine occurrence time and place to predict. However, the current predicted results can not meet the site requirements. According to the event of a longwall mine in Poland, the influence of the type and form of input data on the prediction of mine earthquake was studied, and the method of improving the prediction effect was determined. Practice has proved that using neural network technology can predict the risk of mine shocks.