基于BP神经网络的拉斗铲生产能力预测  

Prediction of Dragline's Production Capacity Based on BP Neural Network

作　　者：肖双双[1,2] 李克民[1,2] 李涛[1,2] 马力[1,2] 常治国[1,2] 

机构地区：[1]中国矿业大学矿业工程学院 [2]煤炭资源与安全开采国家重点实验室

出　　处：《金属矿山》2012年第5期41-43,共3页Metal Mine

基　　金："十一五"国家科技支撑计划项目(编号:2006BAB16B00);中央高校基本科研业务费专项资金项目(编号:2010QNA33;2010ZDP01A02)

摘　　要：在分析拉斗铲生产能力影响因素的基础上,选取实动时间、炸药单耗和采装周期3个因素作为BP神经网络的训练参数建立BP神经网络预测模型,并利用MATLAB实现网络的训练和预测。预测结果表明,BP神经网络能够较好地预测拉斗铲的生产能力,预测输出与实际输出之间的相对误差一般小于10%,预测精度达到了工程要求。Based on analyzing the factors of dragline＇s production capacity, the actual working hours, explosives unit consumption and loading cycle were taken as the training parameters of BP neural network. The BP neural network prediction model was build,and network training and prediction were achieved using MATLAB. The results show that BP neural network could effectively predict the production capacity of dragline. The relative error between predictive output and actual output is normally less than 10%, and its precision of prediction meets the project requirements.

关 键 词：BP神经网络 拉斗铲 生产能力 预测 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]