To address the common issues of insufficient interpretability in intelligent anomaly detection for rotating machinery, as well as the inefficiency and practical limitations caused by the reliance on a large number of labeled samples for model pretraining, this paper presents the multi-granularity scanning Extended isolation Forest with adaptive threshold strategy. A multi-granularity scanning based on interquartile range updating is employed to screen representative local features of the input samples, constituting a multi-granularity feature matrix for anomaly detection and improving feature processing efficiency. Additionally, a randomized 2-D anomaly assessment Extended isolation Forest model is constructed, which enhances model interpretability, optimizes data segmentation paths, avoids artifacts in the Gaussian distribution of anomaly scores, and improves detection accuracy and reliability. A clustering-based adaptive threshold strategy is designed to eliminate dependency on training samples and improve the flexibility and adaptability of the model. Experimental results on bearing and gearbox datasets achieve an accuracy of more than 99.22% and an The area under the ROC curve (ROC-AUC) value approximating 1, demonstrating that the proposed model responds quickly and accurately detects anomalous samples across various operating conditions.
QuatriniECostantinoFDi GravioG, et al. Machine learning for anomaly detection and process phase classification to improve safety and maintenance activities. J Manuf Syst2020; 56: 117–132.
2.
DongHFrusqueGZhaoY, et al. NNG-Mix: improving semi-supervised anomaly detection with pseudo-anomaly generation. Arxiv Preprint Arxiv:2311.11961, 2023.
3.
LiXShaoHJiangH, et al. Modified Gaussian convolutional deep belief network and infrared thermal imaging for intelligent fault diagnosis of rotor-bearing system under time-varying speeds. Struct Health Monit2022; 21(2): 339–353.
4.
ZhaoXJiaM.A novel unsupervised deep learning network for intelligent fault diagnosis of rotating machinery. Struct Health Monit2020; 19(6): 1745–1763.
5.
ZhaoXHuYLiuJ, et al. A novel intelligent multicross domain fault diagnosis of servo motor-bearing system based on domain generalized graph convolution autoencoder. Struct Health Monit2025; 24(3): 1485–1499.
6.
YanSShaoHWangJ, et al. LiConvFormer: a lightweight fault diagnosis framework using separable multiscale convolution and broadcast self-attention. Expert Syst Appl2024; 237: 121338.
7.
CarlettiMTerziMSustoGA.Interpretable anomaly detection with DIFFI: depth-based feature importance of isolation forest. Eng Appl Artif Intell2023; 119: 105730.
8.
YanSShaoHWangX, et al. Few-shot class-incremental learning for system-level fault diagnosis of wind turbine. IEEE/ASME Trans Mechatr2024; 99: 1–10.
9.
DhimanHSDebDMuyeenSM, et al. Wind turbine gearbox anomaly detection based on adaptive threshold and twin support vector machines. IEEE Trans Energy Convers2021; 36(4): 3462–3469.
10.
WangJShaoHYanS, et al. C-ECAFormer: a new lightweight fault diagnosis framework towards heavy noise and small samples. Eng Appl Artif Intell2023; 126: 107031.
11.
WuZYangXWeiX, et al. A self-supervised anomaly detection algorithm with interpretability. Expert Syst Appl2024; 237: 121539.
12.
XiaoYShaoHYanS, et al. Domain generalization for rotating machinery fault diagnosis: a survey. Adv Eng Inform2025; 64: 103063.
13.
BritoLCSustoGABritoJN, et al. An explainable artificial intelligence approach for unsupervised fault detection and diagnosis in rotating machinery. Mech Syst Signal Proc2022; 163: 108105.
14.
WuJZhaoZSunC, et al. Fault-attention generative probabilistic adversarial autoencoder for machine anomaly detection. IEEE Trans Indus Inform2020; 16(12): 7479–7488.
15.
DaiJWangJHuangW, et al. Machinery health monitoring based on unsupervised feature learning via generative adversarial networks. IEEE/ASME Trans Mechatr2020; 25(5): 2252–2263.
16.
PlakiasSBoutalisYS.A novel information processing method based on an ensemble of auto-encoders for unsupervised fault detection. Comp Indus2022; 142: 103743.
17.
ChenZLiZWuJ, et al. Deep residual shrinkage relation network for anomaly detection of rotating machines. J Manuf Syst2022; 65: 579–590.
18.
YanHLiuZChenJ, et al. Memory-augmented skip-connected autoencoder for unsupervised anomaly detection of rocket engines with multi-source fusion. ISA Trans2023; 133: 53–65.
19.
YanSShaoHXiaoY, et al. Hybrid robust convolutional autoencoder for unsupervised anomaly detection of machine tools under noises. Robot Comp Integr Manuf2023; 79: 102441.
20.
LiEBediSMelekW.Anomaly detection in three-axis CNC machines using LSTM networks and transfer learning. Int J Adv Manuf Technol2023; 127(11–12): 5185–5198.
21.
LeeYParkCKimN, et al. LSTM-autoencoder based anomaly detection using vibration data of wind turbines. Sensors2024; 24(9): 2833.
22.
KarczmarekPKiersztynAPedryczW, et al. K-means-based isolation forest. Knowl Based Syst2020; 195: 105659.
23.
LinZLiuXColluM.Wind power prediction based on high-frequency SCADA data along with isolation forest and deep learning neural networks. Int J Elect Power Energy Syst2020; 118: 105835.
24.
LiCGuoLGaoH, et al. Similarity-measured isolation forest: anomaly detection method for machine monitoring data. IEEE Trans Instrum Meas2021; 70: 1–12.
25.
UtkinLAgeevAKonstantinovA, et al. Improved anomaly detection by using the attention-based isolation forest. Algorithms2022; 16(1): 19.
26.
TorresPRamalhoACorreiaL. Automatic anomaly detection in vibration analysis based on machine learning algorithms. In: International conference innovation in engineering. Cham: Springer International Publishing, 2022, pp. 13–23.
27.
SartorDBarbariolTSustoGA.Bayesian active learning isolation forest (B-ALIF): a weakly supervised strategy for anomaly detection. Eng Appl Artif Intell2024; 130: 107671.
28.
ChengTDairiAHarrouF, et al. Monitoring influent conditions of wastewater treatment plants by nonlinear data-based techniques. IEEE Access2019; 7: 108827–108837.
29.
HarrouFDairiATaghezouitB, et al. Automatic fault detection in grid-connected photovoltaic systems via variational autoencoder-based monitoring. Energy Convers Manage2024; 314: 118665.
30.
LiuFTTingKMZhouZH.Isolation forest. In: 2008 8th IEEE international conference on data mining, 15–19 December 2008. Washington, DC: IEEE, 2008, pp. 413–422.
31.
LiuFTTingKMZhouZH.Isolation-based anomaly detection. ACM Trans Knowl Discov Data2012; 6(1): 1–39.
32.
LiuTZhouZYangL.Layered isolation forest: a multi-level subspace algorithm for improving isolation forest. Neurocomputing2024; 581: 127525.
33.
XuHPangGWangY, et al. Deep isolation forest for anomaly detection. IEEE Trans Knowl Data Eng2023; 35(12): 12591–12604.
34.
TaoXPengYZhaoF, et al. A parallel algorithm for network traffic anomaly detection based on isolation forest. Int J Distrib Sensor Netw2018; 14(11): 1550147718814471.
35.
WangHLiSSongL, et al. An enhanced intelligent diagnosis method based on multi-sensor image fusion via improved deep learning network. IEEE Trans Instrum Meas2019; 69(6): 2648–2657.
36.
ZouZXieYHuangK, et al. A docker container anomaly monitoring system based on optimized isolation forest. IEEE Trans Cloud Comp2019; 10(1): 134–145.
37.
HaririSKindMCBrunnerRJ.Extended isolation forest. IEEE Trans Knowl Data Eng2019; 33(4): 1479–1489.
38.
MensiABicegoM.Enhanced anomaly scores for isolation forests. Pattern Recognit2021; 120: 108115.
39.
PevnýT.Loda: lightweight on-line detector of anomalies. Mach Learn2016; 102: 275–304.
40.
ZhaoYNasrullahZHryniewickiMK, et al. LSCP: locally selective combination in parallel outlier ensembles. In: Proceedings of the 2019 SIAM international conference on data mining. Society for industrial and applied mathematics, Calgary, AB, Canada, 2–4 May 2019, pp. 585–593.
41.
LiZZhaoYBottaN, et al. COPOD: copula-based outlier detection. In: 2020 IEEE international conference on data mining (ICDM). Washington, DC: IEEE, 2020, pp. 1118–1123.
QianCJiangQShenY, et al. An intelligent fault diagnosis method for rolling bearings based on feature transfer with improved DenseNet and joint distribution adaptation. Meas Sci Technol2021; 33(2): 025101.
