Unsupervised deep learning-based anomaly detection in mixed gases using sensor array
7 viewsDOI:
https://doi.org/10.54939/1859-1043.j.mst.112.2026.141-148Keywords:
Anomaly detection; Unsupervised learning; Autoencoder; Gas sensor array.Abstract
Anomaly detection in mixed-gas data is a significant challenge for multisensor systems because of their nonlinear characteristics, noise, environmental influences, and scarcity of anomalous data. This study proposes an unsupervised deep-learning anomaly detection method. An unsupervised Autoencoder model was employed to reconstruct two-dimensional image representations of normal mixed-gas data, thereby eliminating the need for anomalous labels during training. To simulate realistic abnormal conditions, anomalous samples were generated using diverse and physically meaningful perturbation transformation methods. An adaptive threshold based on the reconstruction error was used as the criterion for distinguishing between normal and anomalous data points. The effectiveness of the proposed method was evaluated using accuracy and F1-score metrics, which achieved values of 0.97 and 0.92, respectively. The experimental results demonstrate that the proposed approach can effectively learn the characteristics of normal data and accurately detect anomalies in mixed gas datasets. This method improves the safety, reliability, and intelligent monitoring capabilities of gas-sensor array systems in practical applications.
References
[1]. F. Zhang, Z. Zhu, J. Liu, Y. Zhang, M. Xu, and P. Jia, "A Novel Concentration Prediction Technique of Carbon Monoxide (CO) Based on Beluga Whale Optimization-Extreme Gradient Boosting (BWO-XGBoost)", Journal of the Taiwan Institute of Chemical Engineers, Vol. 171, p. 106045, (2025). DOI: https://doi.org/10.1016/j.jtice.2025.106045
[2]. G. Sberveglieri, G. Greco, D. Genzardi, E. Núñez-Carmona, S. Pezzottini, and V. Sberveglieri, "The Electronic Nose: Review on Sensor Arrays and Future Perspectives", Chemical Engineering Transactions, Vol. 95, pp. 265–270, (2022).
[3]. Y. Heng, Y. Zhou, D. H. Nguyen, V. D. Nguyen, and M. Jiao, "An Electronic Nose Drift Compensation Algorithm Based on Semi-Supervised Adversarial Domain Adaptive Convolutional Neural Network", Sensors and Actuators B: Chemical, Vol. 422, p. 136642, (2025). DOI: https://doi.org/10.1016/j.snb.2024.136642
[4]. K. A. Alaghbari, H. S. Lim, M. H. M. Saad, and Y. S. Yong, "Deep Autoencoder-Based Integrated Model for Anomaly Detection and Efficient Feature Extraction in IoT Networks", Internet of Things, Vol. 4, No. 3, pp. 345–365, (2023). DOI: https://doi.org/10.3390/iot4030016
[5]. A. Mikhailova, N. M. Adams, C. A. Hallsworth, F. D. H. Lau, and D. N. Jones, "Unsupervised Deep-Learning-Powered Anomaly Detection for Instrumented Infrastructure", Proceedings of the Institution of Civil Engineers – Smart Infrastructure and Construction, Vol. 172, No. 4, pp. 135–147, (2020).
[6]. W. Ni et al., "Multi-Task Deep Learning Model for Quantitative Volatile Organic Compounds Analysis by Feature Fusion of Electronic Nose Sensing", Sensors and Actuators B: Chemical, Vol. 417, (2024). DOI: https://doi.org/10.1016/j.snb.2024.136206
[7]. J. Fonollosa, S. Sheik, R. Huerta, and S. Marco, "Reservoir Computing Compensates Slow Response of Chemosensor Arrays Exposed to Fast Varying Gas Concentrations in Continuous Monitoring", Sensors and Actuators B: Chemical, Vol. 215, pp. 618–629, (2015). DOI: https://doi.org/10.1016/j.snb.2015.03.028
[8]. X. Zhao et al., "Mixture Gases Classification Based on Multi-Label One-Dimensional Deep Convolutional Neural Network", IEEE Access, Vol. 7, pp. 12630–12637, (2019). DOI: https://doi.org/10.1109/ACCESS.2019.2892754
[9]. S. Zhai, Z. Li, H. Zhang, L. Wang, S. Duan, and J. Yan, "A Multilevel Interleaved Group Attention-Based Convolutional Network for Gas Detection via an Electronic Nose System", Engineering Applications of Artificial Intelligence, Vol. 133, p. 108038, (2024). DOI: https://doi.org/10.1016/j.engappai.2024.108038
