Fault detection with sensor fusion using intelligent immune system

In modern complex systems comprising multiple sensors and other components, fault detection and isolation appears to be a challenging task. An unknown fault in the system can lead to catastrophes. Performance accuracy can be achieved while eliminating the uncertainties and limitations of single sensors through a process of multi-sensor fusion. Inspired by the danger theory of artificial immune systems, an efficient and robust fault detection and isolation scheme is proposed in this paper. The danger theory overcomes such issues as scaling problems, high false positive rates and complexity issues. Furthermore, the characteristics of dendritic cells; i.e., threat detection and information fusion, is utilized for the purpose of fault detection, isolation and sensor fusion. The developed methodology is tested on two cases: the redundant case comprising two infrared sensors and an ultrasonic sensor, and the heterogeneous case comprising two current sensors and a vibration sensor.

[1]  M. Tahir Khan,et al.  Fault detection in robot sensors using negative selection algorithm , 2014, 2014 9th International Conference on Computer Science & Education.

[2]  Reza Langari,et al.  A Hybrid Real-Time System for Fault Detection and Sensor Fusion Based on Conventional Fuzzy Clustering Approach , 2005, The 14th IEEE International Conference on Fuzzy Systems, 2005. FUZZ '05..

[3]  P. Matzinger.  Tolerance, danger, and the extended family. , 1994, Annual review of immunology.

[4]  N. Mort,et al.  Multi-sensor data fusion architecture based on adaptive Kalman filters and fuzzy logic performance assessment , 2002, Proceedings of the Fifth International Conference on Information Fusion. FUSION 2002. (IEEE Cat.No.02EX5997).

[5]  Clarence W. de Silva,et al.  Fault Diagnosis of an Industrial Machine through Sensor Fusion , 2008, Int. J. Inf. Acquis..

[6]  Henri Pierreval,et al.  Fault detection, diagnosis and recovery using Artificial Immune Systems: A review , 2015, Eng. Appl. Artif. Intell..

[7]  P. Matzinger.  The Danger Model: A Renewed Sense of Self , 2002, Science.

[8]  Yu Zhang,et al.  Applied sensor fault detection and validation using transposed input data PCA and ANNs , 2012, 2012 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI).

[9]  Jonathan Timmis,et al.  Application Areas of AIS: The Past, The Present and The Future , 2005, ICARIS.

[10]  Fernando José Von Zuben,et al.  Fault Detection Algorithm for Telephone Systems Based on the Danger Theory , 2005, ICARIS.

[11]  Julie Greensmith,et al.  Introducing Dendritic Cells as a Novel Immune-Inspired Algorithm for Anomoly Detection , 2005, ICARIS.

[12]  Uwe Aickelin,et al.  Danger Theory: The Link between AIS and IDS? , 2003, ICARIS.

[13]  Jonathan Timmis,et al.  Anomaly detection inspired by immune network theory: A proposal , 2009, 2009 IEEE Congress on Evolutionary Computation.

[14]  Muhammad Tahir Khan,et al.  AN OPTIMAL NEURAL NETWORK BASED CLASSIFICATION TECHNIQUE FOR BREAST CANCER DETECTION , 2016 .

[15]  Ren C. Luo,et al.  Multisensor Fusion and Integration: A Review on Approaches and Its Applications in Mechatronics , 2012, IEEE Transactions on Industrial Informatics.

[16]  Federico Castanedo,et al.  A Review of Data Fusion Techniques , 2013, TheScientificWorldJournal.

[17]  Clarence W. de Silva,et al.  Sensors and Actuators: Engineering System Instrumentation, Second Edition , 2015 .

[18]  Simon Coupland,et al.  Fuzzy data fusion for fault detection in Wireless Sensor Networks , 2010, 2010 UK Workshop on Computational Intelligence (UKCI).