A Neural Network Ensemble-Based Approach to Sensor Fault Detection

Abstract:

This paper focuses on the problem of detecting sensor faults in feedback control systems with multistage RBF neural network ensemble-based estimators. The sensor fault detection framework is introduced. The modeling process of the estimator is presented. Fault detection is accomplished by evaluating residuals, which are the differences between the actual values of sensor outputs and the estimated values. The particular feature of the fault detection approach is using the data sequences of multi-sensor readings and controller outputs to establish the bank of estimators and fault-sensitive detectors. A detectability study has also been done with the additive type of sensor faults. The effectiveness of the proposed approach is demonstrated by means of three tank system experiment results.

• Periodical: Key Engineering Materials (Volumes 467-469)
• Pages: 923-927
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.467-469.923
• Online since: February 2011
• Authors: Ai She Shui, Wei Min Chen, Li Chuan Liu, Yong Hong Shui
• Keywords: Fault Detection, Feedback Control Systems, Neural Network Ensemble, Sensor Faults

References

[1] V. Venkatasubramanian, R. Rengaswamy, K. Yin and S. N. Kavuri, A review of process fault detection and diagnosis PartⅠ: quantitative model-based methods, Computer and Chemical Engineering, Vol. 27, No. 3, pp.293-311, (2003).

DOI: 10.1016/s0098-1354(02)00160-6

[2] S. C. Tan, C. P. Lim and M. V. C. Rao, A hybrid neural network model for rule generation and its application to process fault detection and diagnosis, Engineering Applications of Artificial Intelligence, Vol. 20, No. 2, pp.203-213, (2007).

DOI: 10.1016/j.engappai.2006.06.007

[3] V. Venkatasubramanian, R. Rengaswamy, S. N. Kavuri and K. Yin, A review of process fault detection and diagnosis Part III: Process history based methods, Computers and Chemical Engineering, Vol. 27, No. 3, pp.327-346, (2003).

DOI: 10.1016/s0098-1354(02)00162-x

[4] L. K. Hansen and P. Salamon, Neural network ensembles, IEEE Trans Pattern Analysis and Machine Intelligence, Vol. 2, No. 10, pp.993-1001, (1990).

DOI: 10.1109/34.58871

[5] L. Yu, K. K. Lai and S. Y. Wang. Multistage RBF neural network ensemble learning for exchange rates forecasting, Neurocomputing, Vol. 71, No. 16-18, pp.3295-3302, (2008).

DOI: 10.1016/j.neucom.2008.04.029

[6] Y. S. Kim, W. N. Street and F. Menczer, Optimal ensemble construction via meta-evolutionary ensembles, Expert Systems with Applications, Vo30, No. 4, pp.705-714, (2006).

DOI: 10.1016/j.eswa.2005.07.030

[7] R. Isermann, Model-based fault-detection and diagnosis - status and applications, Annual Reviews in Control, Vol. 29, No. 1, 71-85, (2005).

DOI: 10.1016/j.arcontrol.2004.12.002

[8] G. Heredia, A. Ollero, M. Bejar, R. Mahtani. Sensor and actuator fault detection in small autonomous helicopters, Mechatronics, Vol. 18, No. 2, pp.90-99, (2008).

DOI: 10.1016/j.mechatronics.2007.09.007