露天矿高陡边坡变形预报的最优LMBP算法研究

Reasearch of Optimized LMBP Algorithm in Open Pit Mine High and Steep Slope Deformation Prediction

为提高BP神经网络在露天矿高陡边坡变形监测数据预测的精度与可靠性,建立了基于LM算法改进的LMBP神经网络预测模型。以某露天矿边坡监测数据为样本,构建了LMBP最优网络拓扑结构,通过MATLAB编制程序进行了网络训练和预测,应用结果表明:LMBP神经网络具有良好的函数逼近能力及较快的网络收敛能力,且该模型计算结果较为精确,预测精度较高。

To improve the accuracy and reliability of forecasting of high and steep open pit slope deformation monitoring data, puts forward an predictive model of high and steep open pit slope deformation according to the LMBP algorithm.The monitored data in a open-pit mining high and steep slope deformation are viewed as samples, and they are modeled and predicted by MATLAB program. Case study indicates that LMBP neural network had the more nimble effective approximation of function ability and strong convergence ability, and the prediction results are in good agreement with practical deformation and the LMBP model is reliable.