基于机器学习的刮板输送机中部槽磨损预测方法  

Prediction Method for Middle Slot Wear of Scraper Conveyor Based on Machine Learning

作　　者：杨俊叶 申冰 YANG Junye;SHEN Bing(Shijiazhuang Institute of Technology,Shijiazhuang 050228,China)

机构地区：[1]石家庄理工职业学院,石家庄050228

出　　处：《煤炭技术》2023年第4期205-208,共4页Coal Technology

基　　金：2021年度石家庄理工职业学院双高建设专项研究课题(LGSG2021006)。

摘　　要：针对传统刮板输送机中部槽磨损预测方法较为复杂、预测准确率低等问题,结合机器学习理论,提出了一种基于PSO-CNN的刮板输送机中部槽磨损预测方法。通过对中部槽样本数据的处理,构建适用于磨损预测的卷积神经网络(CNN)结构,利用粒子群算法(PSO)对CNN的权值进行评估寻优,避免网络陷入局部最优。实验结果表明:PSO-CNN模型在多种评价指标上均有良好的表现,泛化能力强,性能较为突出;PSO-CNN预测值与试验值的变化曲线贴合度较为一致,预测准确率高,满足实际需求。Based on machine learning theory,a new wear prediction method for middle groove of scraper conveyor based on PSO-CNN is presented,because the traditional prediction method for middle groove of scraper conveyor is complex and has low accuracy.The convolutional neural networks(CNN)structure suitable for wear prediction is constructed by processing the sample data of the middle trough.Particle swarm optimization(PSO)is used to evaluate and optimize the weight of CNN to avoid the network falling into local optimum.The experimental results show that the PSO-CNN model has good performance on a variety of evaluation indicators,with strong generalization ability and outstanding performance.The PSO-CNN predicted value is in good agreement with the change curve of the test value,and the prediction accuracy is high,which meets the actual needs.

关 键 词：刮板输送机 中部槽 PSO-CNN 磨损预测 

分 类 号：TD528[矿业工程—矿山机电] TH227[机械工程—机械制造及自动化]