Robust detection and accommodation of incipient component and actuator faults in nonlinear distributed processes

Abstract

in Wiley InterScience (www.interscience.wiley.com).A class of nonlinear distributed processes with component and actuator faults is presented. An adaptive detection observer with a time varying threshold is proposed that provides additional robustness with respect to false declarations of faults and minimizes the fault detection time. Additionally, an adaptive diagnostic observer is proposed that is subsequently utilized in an automated control reconfiguration scheme that accommodates the component and actuator faults. An integrated optimal actuator location and fault accommodation scheme is provided in which the actuator locations are chosen in order to provide additional robustness with respect to actuator and component faults. Simulation studies of the Kuramoto-Sivashinsky nonlinear partial differential equation are included to demonstrate the proposed fault detection and accommodation scheme. 2008 American Institute of Chemical Engineers AIChE J, 54: 2651AIChE J, 54: -2662AIChE J, 54: , 2008 Keywords: fault detection, fault accommodation, spatially distributed processes, distributed systems IntroductionActuator, sensor and component failures have too often plagued chemical processes, often leading to deteriorating product quality and potentially dangerous process operation, such as runaway conditions. Motivated by the importance of the aforementioned failures, the issue of fault tolerant and fault accommodating controller design has been an active research topic in the chemical engineering community for open-loop stable and open-loop unstable processes. [1][2][3][4][5][6] However, while there has been extensive research from the control community on fault detection and diagnosis of finite dimensional systems using model-based robust and adaptive control techniques, see for example the books 7-9 and references therein, little work can be found for similar treatment of infinite dimensional systems.Previous work on spatially distributed processes has up until recently focused on the design of nonlinear controllers and robust controllers that are specifically tailored to circumvent the computational requirements associated to the infinite dimensional nature of the mathematical description.10-15 A further complexity of such processes lies in the spatial dependence of the concepts of controllability and observability which prompted a number of researchers to address the important topic of actuator and sensor placement. [16][17][18][19] To address the issue of fault tolerance and using modelbased methods along with an adaptive detection observer a fault scheme was first presented in 20,21 for different types of component and actuator faults in general infinite dimensional systems. An adaptive scheme along with robust modifications allowed for the online diagnosis of the component fault parameters. Closer to this work, the concept of a time-varying threshold that minimized the fault detection time was utilized in. 22 Actuator faults with fault tolerant controller design were considered in. [23][24][25][26][27] T...

Actuator Fault Detection and Accommodation in Distributed Parameter Systems

“…Over the recent years, particularly Demetriou [11,12], Armaou [13], and EI-Frarra et al [9,10] studied actuator faulttolerant control in DPSs. Demetriou [11] considered an infinite dimensional adaptive detection observer, which was used to generate a residual signal to detect fault occurrence and assist with fault accommodation for DPSs.…”

“…Demetriou [11] considered an infinite dimensional adaptive detection observer, which was used to generate a residual signal to detect fault occurrence and assist with fault accommodation for DPSs. Armoaou [13] addressed a class of DPSs with component and actuator faults. An adaptive detection observer with a time-varying threshold and an adaptive diagnostic observer were proposed that provide fault declaration and estimation.…”

Fault-tolerant control of sampled-data nonlinear distributed parameter systems

“…Compared with the extensive body of literature on fault diagnosis and fault-tolerant control for lumped parameter systems described by ordinary differential equations (e.g., see [1], [2], [3]), results for spatially-distributed systems have been limited (e.g., see [4], [5], [6]). Major bottlenecks in the design of fault-tolerant control systems for distributed parameter systems include the infinite-dimensional nature of these systems, as well as their complex dynamics characterized by nonlinearities and uncertainties.…”

Safe‐steering of batch process systems

“…Refs. [16][17][18][19][20][21][22]. More recently, the control of non-linear (continuous) processes subject to input constraints and faults that preclude the possibility of operation at the nominal equilibrium point during a fault has been studied.…”