Comparison of two model based residual generation schemes for the purpose of fault detection and isolation applied to a pneumatic actuation system

Abstract

• This is a conference paper, it was presented at SafeProcess Abstract: This paper discusses research carried-out on the development and validation (on a real plant) of a parity-equation and Kalman filter based fault detection and isolation (FDI) system for a pneumatic actuator. The parity and Kalman filter equations are formulated and used to generate residuals that, in turn, are analysed to determine whether faults are present in the system. Details of the design process are given and the experimental results are compared. The results demonstrate that both approaches can successfully detect and isolate faults associated with the sensors, actuators (servo-valves and piping) and the pneumatic cylinder itself. The work is part of a BAE SYSTEMS sponsored project to demonstrate advanced control and diagnosis concepts on an industrial application.

Model-based fault detection and control design – Applied to a pneumatic industrial application

“…Using the description of the system shown in Fig.4 the following relationships (equations) can be derived. (Grewal et al, 2009). …”