基于深度信念网络的磷石膏充填材料强度预测
DOI:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TP3

基金项目:


Strength Prediction of Phosphogypsum Filling material Based on Improved Deep Belief Network
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对目前现有的强度预测方法精度低,文章提出提取输入参数的深层连接的深度信念网络(DBN)强度预测模型,并采用量子粒子群优化算法(Quantum Particle Swarm Optimization,QPSO)来确定DBN的隐含层节点个数和学习率。为获得最优的预测性能,以充填材料的成分及其尺寸作为基于DBN的预测模型的输入,输出充填材料的抗压强度。实验结果显示,该预测方法仅用了1.89s的预测时间且精度达到99.84%,相比于广泛应用的BP神经网络,RVM(Relevance vector machine),SVM(Support Vector Machine)三种算法在精度和时间上都有显著提升。

    Abstract:

    This paper aims to propose a novel approach that can predicts the paste filling material strength of different components with high speed and accuracy because of the accuracy of the existing strength prediction methods is low .To address these problems, the deep belief network (DBN) which can extract the deep layer connection of input parameters is employed to establish the prediction model. In order to realize the optimum prediction performance, the quantum particle swarm optimization algorithm (QPSO) is employed to determine the number of hidden layer nodes and learning rate of DBN. The components of filling material and their dimension are employed as the input of prediction model based on DBN, the output result is the compressive strength of filling material. The experimental results comparing BP neural network ,RVM(Relevance vector machine,RVM)and support vector machine (SVM) show that prediction approach can obtain Significant improvements in prediction time and precision of the 1.89 s by 99.84%.

    参考文献
    相似文献
    引证文献
引用本文

张安安,邓芳明. 基于深度信念网络的磷石膏充填材料强度预测[J]. 科学技术与工程, 2020, 20(18): 7220-7225.
Zhang Anan, Deng Fangming. Strength Prediction of Phosphogypsum Filling material Based on Improved Deep Belief Network[J]. Science Technology and Engineering,2020,20(18):7220-7225.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2019-09-22
  • 最后修改日期:2020-05-01
  • 录用日期:2019-12-19
  • 在线发布日期: 2020-07-28
  • 出版日期:
×
律回春渐,新元肇启|《科学技术与工程》编辑部恭祝新岁!
亟待确认版面费归属稿件,敬请作者关注