The research is focused on the integration of control and fault diagnosis for the purpose of fault-tolerant control. The fault diagnosis activities are concentrated around estimation of unknown faults along with the determination of their location in the diagnosed system. This necessitates the design of integrated and cooperating diagnostic schemes for the process, actuators and sensors. In particular, the research focuses on robust diagnostic schemes based on analytical and soft computing techniques. The resulting solutions incorporate simultaneous state and fault estimation using both robust and adaptive observers.

Another research direction is oriented towards control schemes that can operate in both fault-free and faulty systems. It is realized through the integration of fault estimation and control while taking into account such unappealing effects as disturbances, noise, model uncertainty, estimation errors, as well as constraints imposed on the control and state variables. In particular, the resulting hybrid techniques are based on robust and predictive control. Such a strategy makes it possible to attain robustness to noise and disturbances as well as model uncertainty (robust control) while taking into account time-varying system constraints (predictive control).

The above solutions are developed for nonlinear systems described with analytical models, neural networks, fuzzy logic (Takagi-Sugeno), as well as linear parameter-varying models. The research on robust predictive fault-tolerant control is also conducted for production processes described by discrete event systems, which are modelled with the interval max-plus algebra.