论文部分内容阅读
对海底金矿床开采过程中不同高度岩层位移进行了监测,对岩层变形时间序列重构相空间,用混沌理论揭示了不同高度岩层位移在相空间中的相点距离演变规律。用神经网络建立了岩层变形相空间相点距离演化预测模型,预测了新立矿区海底开采岩层变形,并建立了海底开采岩层变形安全预警系统。采用梯度下降法与混沌优化方法相结合方法训练神经网络,使神经网络预测模型实现快速训练的同时,避免陷入局部极小,同时提高了模型计算精度。研究表明,岩层变形表现出混沌特征,对其相空间重构后,岩层变形的细微变化特征被放大,其内在规律能得到充分展示,为建立海下开采安全预警系统提供了基础。
The displacements of rock formations at different heights during the mining process of the gold seabed gold deposit are monitored, and the phase space is reconstructed for the deformed time series of rock strata. The evolution of the phase distance of phase displacements at different altitudes in the phase space is revealed by chaos theory. The neural network was used to establish the model of predicting the evolution of phase point distance in the deformed phase space of rock strata. The deformation of submarine mining strata in Xinli mining area was predicted and the deformation and safety warning system of submarine mining strata was established. The method of gradient descent and chaos optimization is used to train the neural network so that the neural network predictive model can achieve fast training while avoiding the local minimum and improve the accuracy of the model. The study shows that the deformation of rock strata shows chaotic characteristics. After its phase space is reconstructed, the subtle variations of rock deformation are magnified and their inherent laws can be fully demonstrated, which provides the foundation for the establishment of safety early warning system for mining.