Uncertainty management in continuous systems

In the context of continuous systems, DISCO is developing research into uncertainty management for fault estimation and diagnosis.


In the context of continuous systems, research on uncertainty focuses on the use and improvement of interval-analytic set filtering techniques (guaranteed frames on states and parameters) to solve estimation problems in dynamic systems, while ensuring computational efficiency suitable for real-time applications.

  • Development of set-membership filtering techniques using interval analysis (guaranteed frames on states and parameters), interval Kalman filter (IKF) with optimization of upper bounds on error covariance matrices [hal-04070189], generic interval filter for nonlinear continuous-time systems [hal-04845940], optimally bounded interval Kalman filter (OUBIKF) [hal-03300095].
  • Fault diagnosis in uncertain systems [hal-03689596], particularly for automotive systems [hal-03363452] and aeronautics[hal-03685779]
  • Multi-agent localization in the presence of uncertain observational data [hal-04846108]. Exploitation of zonotopes, ellipsoids...
  • Diagnosability analysis in uncertain systems  [hal-04703936]