Fault detection using sliding mode multiobserver for nonlinear systems: Validation on a real chemical process

Abstract

The design of an accurate observer is still a challenging problem for nonlinear systems subject to fault signals. This paper addresses the problem of simultaneous state and sensor fault estimation for discrete-time nonlinear systems. An uncoupled multimodel approach is adopted to deal with nonlinear systems subject to sensor faults. A simple sensor fault dynamic is exploited to transform this fault signal into a pseudo-actuator fault scenario. A new sliding mode multiobserver is proposed for the estimation of sensor faults and system states. This multiobserver is computed as a fusion of a set of local sliding mode observers with respective activation functions. The asymptotic stability of the proposed multiobserver is formulated in terms of linear matrix inequalities using the Lyapunov theory. A simulation example is retained to demonstrate the effectiveness of the proposed strategy. As well, a real-time application on a transesterification reactor is carried out to validate the theoretical results.

Indirect Measurement of Variables in a Heterogeneous Reaction for Biodiesel Production

“…In addition, no undesirable oscillations are observed with this observer, as found in another report [19]. In addition, the performance index is comparable with that obtained for multiobservers [9,22,23]. It is important to remember that heterogeneous transesterification is considered in this work instead of the homogeneous one considered in the other reported works.…”

“…The methodology of implementing an extended Luenberger observer to estimate the concentration of TG, M, E, and G throughout the transesterification reaction may be the easiest one and offers a balance between precision and complexity. In comparison with other methods such as the EKF [5,10,38], fuzzy methods [19,20], and multimodel methods [22,23], the computational load of Luenberger observers is lower since no recurrent calculations are required. The reported observers consider six [5,10,20] or four [19,21] differential equations; multiobservers consider two equations or set of models with two equations [9,22], but these are reduced models, and only two variables are estimated.…”

“…An evolutionary algorithm was used to minimize the estimation error in an algorithm devoted to estimating uncertain parameters; this algorithm was integrated into an automatic control strategy [21]. An unknown input observer [22] and a sliding-mode observer [23] for uncoupled multimodel representations of transesterification reactors were created; local observers were designed, and interpolation was used to recover the global dynamics. These kinds of observers allow us to consider uncertainties in the studied processes, but the design and tuning can become complex.…”