Robotic-based nonlinear device fault detection with sensor fault and limited capacity for communication

• Published in: Journal of ambient intelligence and humanized computing Vol. 11; no. 12; pp. 6373 - 6385
• Main Authors: Janarthanan, R., Doss, Srinath, Balamurali, R.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-12-2020; Springer Nature B.V
• Subjects: Artificial Intelligence; Asymptotic properties; Communications systems; Computational Intelligence; Engineering; Fault detection; Feedback loops; Fuzzy sets; Fuzzy systems; Linear matrix inequalities; Mathematical analysis; Nonlinear filters; Nonlinearity; Original Research; Robotics; Robotics and Automation; Sensors; Software; User Interfaces and Human Computer Interaction; Wireless networks; Wireless sensor networks; Interval type-2; Fault detection; Robotics; Sensor
• ISSN: 1868-5137; 1868-5145
• DOI: 10.1007/s12652-020-01946-8

This paper proposes the fault detection filter for wireless sensor network. The problem of time-varying nonlinear filter is causing randomly varying nonlinearity due to environment, quantization error and packet dropout are solved by combining Bernoulli distributed white sequence and Robotics based Type-2 T–S (Takagi–Sugeno) Fuzzy method by taking output from sensor and neighbor sensor by solving recursive linear matrix inequalities. The random link failure and uncertainty caused by measurement missing and stochastic communication failure so it can be solved by Interval Type-2 T–S Fuzzy Method Systems (IT-2 T–S FMS) with robotics by closing the feedback loop and Lypunov theory asymptotically stable and satisfying average performance level. The filters are designed for a total defect detection system with a robust mean-square asymptotic stability. So these filters are used in the truck–trailer system for practical output.