Objective In coal mining, anchor digging machines operate under harsh and complex conditions. They are subjected to high-intensity loads, resulting in frequent failures, and the challenge of effective health management becomes increasingly prominent. Therefore, it is crucial to monitor the operational status of anchor digging machines and ensure their reliable performance. This study integrates deep learning and 3D visualization technology, designs a comprehensive health monitoring system framework, and proposes a health monitoring method for anchor digging machines. Methods Based on deep learning, a data-driven remaining useful life (RUL) prediction method was proposed for key components of anchor digging machines. A dual time-length transformer (DT) RUL prediction model incorporating a stacked denoising autoencoder (SDAE) was constructed (SDAE‒DT Net), and particle swarm optimization (PSO) was used for model hyperparameter optimization. The model possessed the structure of dual time-length encoding, which meant that the features of long-time sequences were retained and allowed efficient processing of tightly connected short-time series data. SDAE improved the model, which accurately predicted the RUL in the presence of significant noise interference in the dataset. Experimental validation was conducted using the actual production dataset sourced from the coal mine and the Intelligent Maintenance System (IMS) dataset. The results showed that the SDAE‒DT Net model achieved the highest accuracy and the best prediction performance. On this basis, a 3D visualization health condition monitoring method of the anchor digging machine with data interaction was proposed. The 3D model of the anchor digging machine and the coal mining geological model were constructed using 3D visualization modeling technology. Finally, combined with examples, the 3D visualization health condition monitoring system of the anchor digging machine was developed, which realized the data mapping between the integrated coal mining working face and the 3D visualization model and verified the correctness and feasibility of the method proposed in this study. Results and Discussions For the performance validation experiments of the SDAE‒DT Net model, the sensitivity analysis experiments for PSO hyperparameter optimization showed that the best results were achieved when the population size and inertia factor were 40 and 0.5, respectively. The optimal hyperparameters were: the number of iterations was 239, the number of coder/decoder layers was [4,4], the number of training samples was 153, and the number of hidden neurons was 107. At this point, the PSO‒SDAE‒DT Net model reached the optimal value of each evaluation index in the training set as 0.157, 0.899, 0.192, and 0.087. The ablation experiments explored the effects of the improvements on the model through DT and SDAE. The results showed that the SDAE‒DT Net model was significantly more stable during the training process than the other three experimental sets due to its ability to capture deep features and suppress noise, with a loss of 0.087. Comparison experiments of the prediction results with commonly used models, such as BiGRU, LSTM, and BiLSTM, similarly demonstrated the superiority of the proposed method. Compared with multiple existing methods, experiments were conducted using the IMS dataset. The results showed that the SDAE‒DT Net model has an RMSE value of 0.056 and the best prediction error of 48.36 min, which was the best performance among the models compared. The prediction time of the proposed model was 32 seconds, which was greater than the SVM model's 28 seconds, but the RMSE value was less than 1.214, so both models have their advantages. As a result, the proposed model has the smallest prediction error and higher prediction accuracy. The developed three-dimensional visualization health monitoring system of the anchor digging machine can display real-time environmental data and the operating status of the digging work. In actual production, the real-time monitoring operation status of the anchor digging machine was selected from the system at a specific moment, and the normal value of the operation indices was compared to the actual value. The results showed that all indicators were within the normal range. The vibration signals of the cut-off boom bearing were collected online for prediction to verify the real-time RUL prediction effect of the SDAE‒DT Net model. The results showed that the RMSE value during real-time prediction was 0.098, and the time required was 36 seconds. The predicted RUL value of the 85th sample was 917 min, and the true value was 1 009 min, with a prediction error of 2.94%. The experimental results were close to the results of the historical data. Conclusions The designed health condition monitoring framework for anchor digging machines is capable of real-time and accurate monitoring. The constructed SDAE‒DT Net model effectively integrates features from sequences of varying time lengths within the data, enhancing the RUL prediction accuracy even under noise interference. Utilizing cloud data storage and processing, the developed three-dimensional visualized health condition monitoring system for anchor digging machines enables data interaction and real-time monitoring. The proposed method supports the monitoring of operating status and coal mining anomalies and can serve as a theoretical basis for the efficient operation and maintenance of coal mining equipment.
式(9)~(12)中:为数据集序列;为编码器输入; E、Eposition 分别为线性映射与位置编码;为第i个降维编码器经多头注意力机制与规范化的输出;为降维编码层第i个降维编码器的输出特征;为第i个降维编码器随机丢弃与加入噪声后的损坏数据;为降维编码层的参数集合,, w 为连接权重; b 为偏置向量;为第i个降维编码结构的输出序列;为残差与规范化;为激活函数。
式中,为绝对值, S 为最后一个升维编码器的输出。堆叠多层DAE是将前一层DAE经编码函数得到的输出作为后一层的输入,后一层经过类似前一层的方式再次重构数据。维度转换层堆叠3层DAE同时训练,参数集合与使用梯度下降法调整所有连接权重和偏置向量。选择梯度下降法的原因是其能够高效处理高维参数空间,且适用于DAE的非线性特征学习需求,从而保证模型在复杂特征表示上的训练效果。在重构误差最小时训练完成,此时得到经过降噪处理的特征数据。模型优化的目标共包含预测误差、重构误差两部分。预测误差通常选择平均绝对误差(MAE)作为损失值,总优化目标为:
依据动态数据交互的掘锚机3维模型构建方法构建掘锚机虚拟模型,通过数据驱动的掘锚机剩余寿命预测方法实现截割大臂轴承RUL预测,并结合云端数据存储与处理方法进行实时的可视化健康状态监测。使用VUE前端开发框架与Django后端深度学习模型集成训练框架,通过Visual Studio Code软件开发基于B/S架构的掘锚机3维可视化健康状态监测系统,如图13所示。系统采用3维模型展示的形式,对掘锚机的位置、运行环境、航偏角度以及报警信息等数据可视化展示,并实时监控其关键零部件的RUL值。通过点击3维模型的方式,即可获得实时剩余寿命预测曲线,并对处于寿命末期的设备进行告警。
WangGuofa, ZhangLiang, LiShoubin,et al.Progresses in theory and technological development of unmanned smart mining system[J].Journal of China Coal Society,20,48(1):34‒53.
WangGuofa, DuYibo.Development direction of intelligent coal mine and intelligent mining technology[J].Coal Science and Technology,2019,47(1):1‒10. doi:10.13199/j.cnki.cst.2019.01.001
YangXueqi, GaoXinqin, ZhengHaiyang.Operation parameters optimization method of coal flow transportation equipment based on convolutional neural network[J].Mining,Metallurgy & Exploration,2024,41(4):1793‒1806. doi:10.1007/s42461-024-01031-0
[6]
BaoJiusheng, ZhangQuanli, GeShirong,et al.Basic research and application practice of unmanned auxiliary transportation system in coal mine[J].Journal of China Coal Society,2023,48(2):1085‒1098. doi:10.13225/j.cnki.jccs.2022.1600
GaoXinqin, YangXueqi, ZhengHaiyang.Residual life prediction method of belt conveyors based on MDT learning[J].China Mechanical Engineering,2024,35(8):1435‒1448.
LiaoLinxia, KöttigF.Review of hybrid prognostics approaches for remaining useful life prediction of engineered systems,and an application to battery life prediction[J].IEEE Transactions on Reliability,2014,63(1):191‒207. doi:10.1109/TR.2014.2299152
[11]
LiTianmei, SiXiaosheng, LiuXiang,et al.Data-model interactive remaining useful life prediction technologies for stochastic degrading devices with big data[J].Acta Automatica Sinica,2022,48(9):2119‒2141.
LiTianmei, PeiHong, SiXiaosheng,et al.Prognosis for stochastic degrading systems with massive data:A data-model interactive perspective[J].Reliability Engineering & System Safety,2023,237:109344. doi:10.1016/j.ress.2023.109344
[14]
HuangZeyi, XuZhengguo, WangWenhai,et al.Remaining useful life prediction for a nonlinear heterogeneous Wiener process model with an adaptive drift[J].IEEE Transactions on Reliability,2015,64(2):687‒700. doi:10.1109/tr.2015.2403433
[15]
ZhangQiang, ZhangJiayao, LvFuyan.Prediction of pick wear degradation states of road header based on Wiener process[J].Journal of Vibration and Shock,20,42(1):207‒214.
PengWeiwen, LiYanfeng, YangYuanjian,et al.Bivariate analysis of incomplete degradation observations based on inverse Gaussian processes and copulas[J].IEEE Transactions on Reliability,2016,65(2):624‒639. doi:10.1109/tr.2015.2513038
[18]
WangHaikun, HuangHongzhong, LiYanfeng,et al.Condition-based maintenance with scheduling threshold and maintenance threshold[J].IEEE Transactions on Reliability,2015,65(2):513‒524. doi:10.1109/tr.2015.2487578
[19]
PeiHong, HuChanghua, SiXiaosheng,et al.Review of machine learning based remaining useful life prediction methods for equipment[J].Journal of Mechanical Engineering,2019,55(8):1‒13. doi:10.3901/jme.2019.08.001
KhanS, YairiT.A review on the application of deep learning in system health management[J].Mechanical Systems and Signal Processing,2018,107:241‒265. doi:10.1016/j.ymssp.2017.11.024
[22]
LiXiang, DingQian, SunJianqiao.Remaining useful life estimation in prognostics using deep convolution neural networks[J].Reliability Engineering & System Safety,2018,172:1‒11. doi:10.1016/j.ress.2017.11.021
[23]
ZhouYu, FangQian, PeiZexuan,et al.Fault diagnosis of ro-lling bearing based on improved sparrow search algorithm optimized LSTM[J].Advanced Engineering Sciences,2024,56(2):289‒298
DengFeiyue, ChenZhe, HaoRujiang,et al.Research on bearing remaining useful life prediction based on an MsTCN‒Transformer model[J].Journal of Vibration and Shock,2024,43(4):279‒287.
JiangXianliang, ChenLingyu, ZhengJieji,et al.High-preci-sion control method of direct drive components based on digital twin model[J].Journal of Mechanical Engineering,2021,57(17):98‒109. doi:10.3901/jme.2021.17.098
YangLinyao, ChenSiyuan, WangXiao,et al.Digital twins and parallel systems:State of the art,comparisons and prospect[J].Acta Automatica Sinica,2019,45(11):2001‒2031.
HazrathosseiniA, Moradi AfrapoliA.The advent of digital twins in surface mining:Its time has finally arrived[J].Resources Policy,2023,80:103155. doi:10.1016/j.resourpol.2022.103155
[37]
ChiHuanlei, YuanZhi, CaoYan,et al.Study on digital twin-based smart fully-mechanized coal mining workface monitoring technology[J].Coal Science and Technology,2021,49(10):153‒161.
DingHua, YangLiangliang, YangZhaojian,et al.Health prediction of shearers driven by digital twin and deep learning[J].China Mechanical Engineering,2020,31(7):815‒823. doi:10.3969/j.issn.1004-132X.2020.07.007
YangChunyu, BuLingchao, ChenBin.Operation optimization method of long-distance belt conveyor driven by digital twins[J].China Industrial Economics,2024,11:2204‒2218.
GeShirong, ZhangFan, WangShibo,et al.Digital twin for smart coal mining workface:Technological frame and construction[J].Journal of China Coal Society,2020,45(6):1925‒1936.
LinShan, DongMiao, CaoX,et al.The pre-trained explainable deep learning model with stacked denoising autoencoders for slope stability analysis[J].Engineering Analysis with Boundary Elements,2024,163:406‒425. doi:10.1016/j.enganabound.2024.03.019
[46]
TijjaniS, Ab‒WahabM N, Mohd‒NoorM H.An enhanced particle swarm optimization with position update for optimal feature selection[J].Expert Systems with Applications,2024,247:123337. doi:10.1016/j.eswa.2024.123337
[47]
LiuShujun, XuTong, DuXiaoze,et al.A hybrid deep learning model based on parallel architecture TCN‒LSTM with Savitzky‒Golay filter for wind power prediction[J].Energy Conversion and Management,2024,302:118122. doi:10.1016/j.enconman.2024.118122
[48]
ShiJiayu, ZhongJingshu, ZhangYuxuan,et al.A dual attention LSTM lightweight model based on exponential smoothing for remaining useful life prediction[J].Reliability Engineering & System Safety,2024,243:109821. doi:10.1016/j.ress.2023.109821
[49]
QianYuning, YanRuqiang, GaoR X.A multi-time scale approach to remaining useful life prediction in rolling bearing[J].Mechanical Systems and Signal Processing,2017,83:549‒567. doi:10.1016/j.ymssp.2016.06.031
[50]
LiuYongbin, HeBing, LiuFang,et al.Remaining useful life prediction of rolling bearings using PSR,JADE,and extreme learning machine[J].Mathematical Problems in Engineering,2016,2016:8623530. doi:10.1155/2016/8623530
[51]
LiaoHaitao, ZhaoWenbiao, GuoHuairui.Predicting remaining useful life of an individual unit using proportional hazards model and logistic regression model[C]//Proceedings of the 2006 Annual Reliability and Maintainability Symposium.Newport Beach:IEEE,2006:127‒132.
[52]
Ben AliJ, Chebel‒MorelloB, SaidiL,et al.Accurate bearing remaining useful life prediction based on Weibull distribution and artificial neural network[J].Mechanical Systems and Signal Processing,2015,56:150‒172. doi:10.1016/j.ymssp.2014.10.014
[53]
YanMingming, WangXingang, WangBingxiang,et al.Bea-ring remaining useful life prediction using support vector machine and hybrid degradation tracking model[J].ISA Transactions,2020,98:471‒482. doi:10.1016/j.isatra.2019.08.058
[54]
GuoLiang, LiNaipeng, JiaFeng,et al.A recurrent neural network based health indicator for remaining useful life prediction of bearings[J].Neurocomputing,2017,240:98‒109. doi:10.1016/j.neucom.2017.02.045
[55]
ChangZihan, YuanWei, HuangKou.Remaining useful life prediction for rolling bearings using multi-layer grid search and LSTM[J].Computers and Electrical Engineering,2022,101:108083. doi:10.1016/j.compeleceng.2022.108083
[56]
GuptaM, WadhvaniR, RasoolA.A real-time adaptive model for bearing fault classification and remaining useful life estimation using deep neural network[J].Knowledge‒Based Systems,2023,259:110070. doi:10.1016/j.knosys.2022.110070