充填管道失效风险性预测精度研究

来源期刊:中南大学学报(自然科学版)2014年第8期

论文作者:张钦礼 周碧辉 王新民 周登辉 王石

文章页码:2805 - 2812

关键词:充填管道,危险性预测;主成分分析法;改进的BP神经网络;预测精度

Key words:backfilling pipeline; risk prediction; principal components analytic method; improved BP neural network; prediction precision

摘    要:为更准确预测矿山充填管道失效风险性,建立主成分分析与改进BP神经网络相结合的评价模型。选取10项评价指标作为充填管道失效风险性的评判指标,统计10个矿山的样本数据,并运用主成分分析法对这10个样本数据进行预处理,得出主要成分,再利用改进的BP神经网络模型进行预测,最终得到更准确的管道失效风险预测结果。研究结果表明,所得到实际预测结果与期望值之间的相对误差分别为2.31%,1.68%,3.02%。预测相对误差控制在4%以内,较未经主成分分析处理的标准BP神经网络预测精度更为准确。利用主成分分析法与改进的BP神经网络相结合建立的充填管道失效评价模型具有分析速度快、预测精度高的特点,为矿山充填管道失效风险预测提供了一种更为完善的方法。

Abstract: In order to predict the invalidation risk of backfilling pipeline more accurately, the evaluation system of principal component analysis and the improved BP neural network was established. Ten primary factors affecting invalidation of backfilling pipeline were considered. And the sample data of ten mines were counted, which were processed by principal component analysis, producing the main ingredients, as the input data of the optimized BP Neural Network which ended up with a more accurate risk prediction on invalidation of backfilling pipeline. The results indicate that the expected value relative errors of the three mines are 2.31%, 1.68%, 3.02%, respectively, all controlled within 4%. The prediction accuracy is improved greatly compared with the prediction by standard BP neural network without principal components analysis. The invalidation risk assessment of backfill pipeline by the establishment of the principal components analytic method and the optimized BP neural network has the advantage of rapid analysis and high precision of prediction. A better evaluation is provided for the mine backfilling pipeline invalidation risk analysis.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号