基于量子神经网络拟合法的矿区地表变形监测

Surface Deformation Monitoring of Mining Area Based on Quantum Neural Network Fitting Method

矿区地表变形监测受到矿区地质构造条件、开采规模、采矿工艺等因素的影响,目前常规的矿区地表变形监测方法具有流程繁琐、工作量大、监测精度低等不足,为此,提出了一种基于量子神经网络拟合法的矿区地表变形监测方法。该方法通过将矿区监测点x、y坐标作为神经网络输入层神经元,将监测点的高程异常量(ξ)作为神经网络的输出层神经元,经多次迭代获得最优解。基于某矿区GPS监测数据,分别采用二次多项式拟合、BP神经网络拟合以及所提方法进行对比试验,并引入内、外符合精度作为各方法拟合精度的评价标准,结果表明:对于不同分布的监测点以及不同数量的监测点,所提方法相对于其余2种方法而言具有较高的内、外符合精度及较小的残差,对于提高矿区变形监测精度有一定的参考价值。

The surface deformation monitoring of mining area is influenced by the differences of geological structure conditions,mining scale and mining methods. The classical surface deformation monitoring methods of mining area with the defects of process trival,heavy workload and low accuracy,so,the surface deformation monitoring method of mining area based on quantum neural network fitting method is proposed. The coordinates（ x,y） of monitoring points in mining area are taken as the input layer neurons of quantum neural network,the abnormal height（ ξ） of the monitoring points in mining area are taken as the output layer neurons of quantum neural network,optimal solutions are obtained by multiple iterations of quantum neural network. Based on the GPS monitoring data of a mining area,the quadratic polynomial fitting method,BP neural network fitling method and the quantum neural network fitting method proposed in this paper are adopted to conduct contrast experiment,the internal precision and external precision are taken as the evaluation standard of the fitting precision of the above three methods. The experimental results show that the internal precision and external precision of quantum neural network fitting method proposed in this paper are higher than the quadratic polynomial fitting method and BP neural network filling method,the residual errors of the quantum neural network fitting method proposed in this paper is lower than the quadratic polynomial fitting method and BP neural network filling method,the experimental results further show that the quantum neural network fitting method proposed in this paper has reference for the improvement of surface deformation monitoring precision of mining area.