Fault detection and identification technique in control systems with redundant inertial measurement unit

Aeronautical and Space-Rocket Engineering

Aeronautical engineering


Gorbatenko S. A.*, Klionovska K. K.**

Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: mai_kaf604@mail.ru
**e-mail: klionovska@gmail.com


Development of aviation technology requires the use of a variety of measuring systems to determine the motion and positioning variables of the aircrafts in the space. Requirements for the reduction of weight and dimensions of such systems are the first since the objects themselves demonstrate a tendency to reduce their weight and size characteristics.

Availability of inertial measurements units (IMU) allows to create tiny strapdown inertial navigation system (SINS), which have small weight and dimensions, and can be used in areas such as attitude control system, radar, avionics, data transmission system, navigation, etc. The work of these systems is accompanied by a variety of errors and failures, which leads to the urgency of early detection and recognition of their errors. This difficult problem can be solved on the basis of different approaches and with varying degrees of success.

The article describes one of the possible ways to solve this problem, such as using combination of two methods like a parity space approach and the discrete wavelet transform in the redundant system with additional IMU. Such combination allows to detect three faulty sensors and to isolate each of them out from the system during the system operation. The one of the positive advantage of this algorithm that it is not demanded to know the dynamic of the system model, except orientation of the sensors in the measurement system organized with redundant IMU. The simulation of the successful operation of this algorithm with two types of sensor faults, such as bias and stuck, are demonstrated in this work.


inertial measurement unit, fault detection and identification, parity space approach, discrete wavelet transformation


  1. Zhang Han Qiu, Yan Yong A wavelet-based approach to abrupt fault detection and diagnosis of sensors, IEEE Transactions on Instrumentation and Measurement, October 2001, vol. 50, no. 5, pp. 1389-1396.
  2. Won Hee Lee, Kwang Hoon Kim, Chan Gook Park and Jang Gyu Lee. Two-Faults Detection and Isolation Using Extended Parity Space Approach, Journal of Electrical Engineering& Technology, 2012, vol. 7, no. 3, pp. 411-419.
  3. Yoon Seungho, Kim Swungkeun, Kim Youdan, Choi Jiyoung, Lee Jangho, Kim Eungtai. Experimental Evaluation of Model-free Hybrid Fault Detection and Isolation, Control Engineering Practice, 2011, vol. 19, no. 2, pp. 158-173.
  4. Kim S., Kim Y. and Park C. Failure diagnosis of skew- configured aircraft inertial sensors using wavelet decomposition. IET Control Theory and Applications, 2007, vol. 1, no. 5, pp. 1390-1397.
  5. Dobeshi I. Desyat lektsii po veivletam, Izhevsk, NITs «Regulyarnaya i khaoticheskaya dinamika», 2001, 464 p.
  6. Dyakonov V.P. MATLAB i SIMULINK dlya radio- inzhenerov (MATLAB and SIMULINK for radio engineers), Moscow, DNK Press, 2011, 976 p.

mai.ru — informational site of MAI

Copyright © 1994-2023 by MAI