Anomaly identification and correction of multi-dimensional operating conditions for a data-driven vehicle noise and vibration platform based on fusion of deep learning models
Restricted accessResearch articleFirst published online 2026
Anomaly identification and correction of multi-dimensional operating conditions for a data-driven vehicle noise and vibration platform based on fusion of deep learning models
With the proliferation of data-driven methods in automotive noise, vibration, and harshness (NVH) analysis, the digital transformation of NVH performance evaluation has become increasingly imperative. However, in the actual testing process, signals are inevitably affected by abnormal interference, resulting in a decrease in the evaluation accuracy of NVH performance. To solve this problem, we propose a new adaptive anomaly detection and correction framework. The key methodological innovation lies in the IResNet–IFCNN collaborative architecture, which introduces an improved residual network (IResNet) for high-precision anomaly identification and an improved fully connected network (IFCNN) for adaptive multi-condition correction. The main contribution is a closed-loop “detection – matching – correction” mechanism, which dynamically selects specific weights based on the type of anomaly. Verified on the real vehicle test data, the correction accuracy reached 93.33%, significantly enhancing the robustness of the intelligent NVH assessment under interference.
ShenXZuoSHanL.Interior vehicle noise quality prediction using support vector machines. J Vib Meas Diagn2011; 31: 55–58.
2.
FangYZhangT.Sound quality investigation and improvement of an electric powertrain for electric vehicles. IEEE Trans Ind Electron2018; 65: 1149–1157.
3.
XuZXiaXLaiS.Improvement of interior sound quality for passenger cars based on optimization of sound pressure distribution in low frequency. Appl Acoust2018; 130: 43–51.
4.
ZhangXCChengJLuJW.Sound quality evaluation of pure electric vehicles with a subjective and objective unified evaluation method. Int J Veh Des2022; 88: 283–303.
5.
SianoDPanzaMABadanJP. Prediction of interior vehicle noise by means of NARX neural networks (No. 2018-01-1538). SAE Technical Paper, 2018.
6.
JoYJeongKAhnS, et al. Data augmentation methods for electric automobile noise design from multi-channel steering accelerometer signals. Lect Notes Netw Syst2023; 542: 679–684.
7.
StenderMHoffmannN. Deep learning for predicting brake squeal. In: International Conference on Noise and Vibration Engineering (ISMA 2020) and International Conference on Uncertainty in Structural Dynamics (USD 2020), 2020, pp. 3327–3337.
8.
HuangZHChenHYZouAH.Study on MSA–SVR method for predicting the sound quality of vehicle door closing. Noise Vib Control2021; 41: 145–150.
9.
PangJMaoTJiaW, et al. Prediction and analysis of vehicle interior road noise based on mechanism and data series modeling. Sound Vib2024; 58: 59–80.
10.
TsokaktsidisDEMaederMVon WysockiT.Generating narrowband automotive acoustic models from measured data with artificial neural networks and Bayesian optimization of hyperparameters. J Theor Comput Acoust2025; 33(2): 2550006.
11.
Morris-KirbyR. The use of “big data” for the analysis and design of vehicle sound packages (No. 2018-01-1570). SAE Technical Paper, 2018.
12.
HuangHBWuJHHuangXR, et al. A generalized inverse cascade method to identify and optimize vehicle interior noise sources. J Sound Vib2020; 467: 115062.
13.
WangYGuoHYangC (eds). Vehicle interior noise mechanism and prediction. In: Vehicle interior sound quality: analysis, evaluation and control. Springer Nature Singapore, 2022, pp. 5–62.
14.
TsokaktsidisDENauCMarburgS. Time domain full vehicle interior noise calculation from component level data by machine learning (No. 2020-01-1564). SAE Technical Paper, 2020.
15.
TsokaktsidisDENauCMaederM, et al. Using rectified linear unit and swish based artificial neural networks to describe noise transfer in a full vehicle context. J Acoust Soc Am2021; 150(3): 2088–2105.
16.
Ramos-RomeroCAsensioC. Asphalt-surface defects detection, based on tyre/road noise analysis and geo-processing. In: Proceedings of the international congress on acoustics, Aachen, Germany, 9–13 September 2019.
17.
SafontGSalazarARodriguezA, et al. Multichannel signal processing for road surface identification. In: ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Virtual conference, 4–9 May 2020, pp. 3052–3056, IEEE.
18.
LeeDKimJCKimM, et al. Intelligent tire sensor-based real-time road surface classification using an artificial neural network. Sensors2021; 21: 3233.
19.
KumarATiwariRGTrivediNK. Hybrid deep learning framework for road surface classification: integrating autoencoder-based denoising and CNN-based classification. In: 2025 international conference on intelligent control, computing and communications (IC3), Mathura, India, 13–14 February 2025, pp. 487–492. IEEE.
20.
MaoRWuGWuJ, et al. A lightweight convolutional neural network for road surface classification under shadow interference. Knowl-Based Syst2024; 306: 112761.
21.
WangTLiuZChenW, et al. A multi-dimensional driving condition recognition method for data-driven vehicle noise and vibration platform based on speed separation strategy and deep learning model. Mech Syst Signal Process2025; 236: 113001.
22.
SongDHongSHaC, et al. Acoustic analysis and data-driven control of vehicle NVH: a framework for manufacturing process optimization. Appl Acoust2025; 233: 110618.
23.
KhakshourniaSHaghighiSSMajidiM, et al. A hybrid strategy of OTPA and machine learning for efficient root-cause analysis of NVH in multi-source systems. Appl Acoust2026; 242: 111060.
24.
SongDHaCJangJ, et al. A study on NVH monitoring automation using AI/DS. In: INTER-NOISE and NOISE-CON congress and conference proceedings. Institute of Noise Control Engineering, 2025, Vol. 272, pp. 2054–2059.
25.
MaYDaiRLiuT, et al. Research on vehicle road noise prediction based on AFW–LSTM. Machines2025; 13(5): 425.
26.
JafarianKMobinMJafari-MarandiR, et al. Misfire and valve clearance faults detection in the combustion engines based on a multi-sensor vibration signal monitoring. Measurement2018; 128: 527–536.
27.
AbadMRAAMoosavianAKhazaeeM. Wavelet transform and least square support vector machine for mechanical fault detection of an alternator using vibration signal. J Low Freq Noise Vib Active Control2016; 35(1): 52–63.
28.
GhaderiHKabiriP.Automobile engine condition monitoring using sound emission. Turk J Electr Eng Comput Sci2017; 25(3): 1807–1816.
29.
MathewSKZhangY.Acoustic-based engine fault diagnosis using WPT, PCA and Bayesian optimization. Appl Sci2020; 10(19): 6890.
30.
RamtekeSMChelladuraiHAmarnathM.Diagnosis and classification of diesel engine components faults using time–frequency and machine learning approach. J Vib Eng Technol2021; 10: 175–192.
31.
Pérez-VázquezAAnzures-GarcíaMSánchez-GálvezLA.Vehicle engine fault diagnosis approach based on a decision tree and knowledge base. Int J Comb Optim Prob Inform2024; 15(2): 185–197.
KemalkarAKBairagiVK. Engine fault diagnosis using sound analysis. In: 2016 International Conference on Automatic Control and Dynamic Optimization Techniques (ICACDOT), Hinjawadi, India, 9–10 September 2016, pp. 943–946. IEEE.
34.
GongCSASuCHSTsengKH. Implementation of machine learning for fault classification on vehicle power transmission system. IEEE Sens J2020; 20(24): 15163–15176.
35.
ChakrapaniGSugumaranV.Transfer learning based fault diagnosis of automobile dry clutch system. Eng Appl Artif. Intell2023; 117: 105522.
36.
SaiSAChakrapaniGAnnamalaiK, et al. Enhanced fault detection in automobile clutch system using CATboost with feature fusion method. Phys Scr2024; 99(9): 095012.
37.
KarthikeyanHLSridharanNVBalajiPA, et al. Diagnosing faults in suspension system using machine learning and feature fusion strategy. Arab J Sci Eng2024; 49(11): 15059–15083.
38.
ArgawalRKalelDHarshitM, et al. Sensor fault detection using machine learning technique for automobile drive applications. In: 2021 National Power Electronics Conference (NPEC), Virtual conference, 15–17 December 2021, pp. 1–6. IEEE.
39.
ZangBGaoHYangX, et al. Fault diagnosis of charging facilities based on improved RNN. J Phys Conf Ser2022; 2301: 012008.
40.
WuW.Automotive motor fault diagnosis model integrating machine learning algorithm and fuzzy control theory. Int J Fuzzy Syst2025; 27: 1–14.
41.
ZhaiSWangZ. Research on automobile fault diagnosis model based on neural network. In: 2024 IEEE 4th International Conference on Power, Electronics and Computer Applications (ICPECA), Shenyang, China, 26–28 January 2024, pp. 411–418. IEEE.
42.
YangCQiaoZLiuL, et al. Positive-incentive noise in artificial intelligence-enabled machine fault diagnosis. Struct Health Monit2025; 2025: 14759217251370358.
43.
HeKZhangXRenS, et al. Deep residual learning for image recognition. ArXiv Prepr. arXiv:1512.03385, 2015, pp. 171–180.
44.
ZhangKTangBDengL, et al. A hybrid attention improved ResNet-based fault diagnosis method of wind turbines gearbox. Measurement2021; 179: 109491.
45.
LuoYWangZ. An improved ResNet algorithm based on CBAM. In: 2021 International Conference on Computer Network, Electronic and Automation (ICCNEA), Xi'an, China, 24–26 September 2021, pp. 121–125. IEEE.
46.
LiangPWangWYuanX, et al. Intelligent fault diagnosis of rolling bearing based on wavelet transform and improved ResNet under noisy labels and environment. Eng Appl Artif Intell2022; 115: 105269.
47.
ZouXWuCLiuH, et al. Improved ResNet-50 model for identifying defects on wood surfaces. Signal Image Video Process2023; 17: 3119–3126.
48.
LiXXuXHeX, et al. Intelligent crack detection method based on GM-ResNet. Sensors2023; 23: 8369.
49.
WenLXiaoZXuX, et al. Disaster recognition and classification based on improved ResNet-50 neural network. Appl Sci2025; 15: 5143.
50.
KhattapMGSallahMDahouA, et al. Advancing multiple sclerosis diagnosis through an innovative hybrid AI framework incorporating multi-view ResNet and quantum RIME-inspired metaheuristics. Ain Shams Eng J2025; 16: 103241.
51.
NisaMLeszczukMJuszkaD, et al. Improved binary classification of underwater images using a modified ResNet-18 model. SSRN J2025; 2025: 5097940.
52.
SahuSKMohapatraDPRoutJK, et al. A LSTM–FCNN based multi-class intrusion detection using scalable framework. Comput Electr Eng2022; 99: 107720.
53.
OuyangJLiTLiQ. Time series prediction with missing values: a bidirectional fully connected neural network approach. In: 2023 35th Chinese Control and Decision Conference (CCDC), Yichang, China, 20–23 May 2023, pp. 2245–2250. IEEE.
54.
BhuiyanMRAbdullahJHashimN, et al. Hajj pilgrimage abnormal crowd movement monitoring using optical flow and FCNN. J Big Data2023; 10(1): 86.
55.
WangXLiuCLiuP, et al. A new method for cross-calibration between LIBS systems based on a domain adaptive fully connected network. J Anal At Spectrom2024; 39(2): 558–572.
56.
DongNZhangTWangR, et al. An evolutionary algorithm based on fully connected weight networks for mixed-variable multi-objective optimization. Inf Sci2024; 659: 120053.
57.
XiaoCZhangSHuY, et al. Robust optimization of geoenergy production using data-driven deep recurrent auto-encoder and fully-connected neural network proxy. Expert Syst Appl2024; 242: 122797.
58.
El IchiAKaddahWEl BouzM, et al. TER–FCN: Tensor Einstein Reduced Fully Connected Network applied to Region-based Convolutional Neural Networks for object detection. Signal Image Video Process2025; 19(8): 1–9.
