Model-based fault-detection and diagnosis – status and applications

https://doi.org/10.1016/j.arcontrol.2004.12.002Get rights and content

Abstract

For the improvement of reliability, safety and efficiency advanced methods of supervision, fault-detection and fault diagnosis become increasingly important for many technical processes. This holds especially for safety related processes like aircraft, trains, automobiles, power plants and chemical plants. The classical approaches are limit or trend checking of some measurable output variables. Because they do not give a deeper insight and usually do not allow a fault diagnosis, model-based methods of fault-detection were developed by using input and output signals and applying dynamic process models. These methods are based, e.g., on parameter estimation, parity equations or state observers. Also signal model approaches were developed. The goal is to generate several symptoms indicating the difference between nominal and faulty status. Based on different symptoms fault diagnosis procedures follow, determining the fault by applying classification or inference methods. This contribution gives a short introduction into the field and shows some applications for an actuator, a passenger car and a combustion engine.

Introduction

Within the automatic control of technical systems, supervisory functions serve to indicate undesired or not permitted process states, and to take appropriate actions in order to maintain the operation and to avoid damage or accidents. The following functions can be distinguished:

  • (a)

    monitoring: measurable variables are checked with regard to tolerances, and alarms are generated for the operator;

  • (b)

    automatic protection: in the case of a dangerous process state, the monitoring function automatically initiates an appropriate counteraction;

  • (c)

    supervision with fault diagnosis: based on measured variables, features are calculated, symptoms are generated via change detection, a fault diagnosis is performed and decisions for counteractions are made.

The big advantage of the classical limit-value based supervision methods (a) and (b) is their simplicity and reliability. However, they are only able to react after a relatively large change of a feature, i.e., after either a large sudden fault or a long-lasting gradually increasing fault. In addition, an in-depth fault diagnosis is usually not possible. Therefore (c) advanced methods of supervision and fault diagnosis are needed which satisfy the following requirements:

  • (1)

    Early detection of small faults with abrupt or incipient time behavior;

  • (2)

    Diagnosis of faults in the actuator, process components or sensors;

  • (3)

    Detection of faults in closed loops;

  • (4)

    Supervision of processes in transient states.

A general survey of supervision, fault-detection and diagnosis methods is given in (Isermann, 1997). In the following model-based fault-detection methods are considered, which allow a deep insight into the process behavior.

Section snippets

Process model-based fault-detection methods

Different approaches for fault-detection using mathematical models have been developed in the last 20 years, see, e.g., (Chen & Patton, 1999; Frank, 1990, Gertler, 1998, Himmelblau, 1978, Isermann, 1984, Isermann, 1997; Patton, Frank, & Clark, 2000; Willsky, 1976). The task consists of the detection of faults in the processes, actuators and sensors by using the dependencies between different measurable signals. These dependencies are expressed by mathematical process models. Fig. 1 shows the

Fault diagnosis methods

The task of fault diagnosis consists of the determination of the type of fault with as many details as possible such as the fault size, location and time of detection. The diagnostic procedure is based on the observed analytical and heuristic symptoms and the heuristic knowledge of the process. The inputs to a knowledge-based fault diagnosis system are all available symptoms as facts and the fault-relevant knowledge about the process, mostly in heuristic form. The symptoms may be presented just

Applications of model- and signal-based fault diagnosis

In the following some results from case studies and in-depth investigations of model-based fault-detection methods are briefly described. The examples are selected such that they show different approaches and process adapted solutions which can be transferred to other similar technical processes.

Conclusion

After a short introduction into model-based fault-detection methods, like parameter estimation, observers and parity equations, and fault-diagnosis methods, like classification and inference methods, three application examples were shown. For a DC motor actuator of an aircraft cabin pressure control the combination of parameter estimation and parity equations allows the detection of several parametric and additive faults by using four measurements, followed by the fault diagnosis with

Rolf Isermann received both the Dipl.-Ing degree in mechanical engineering and the Dr.-Ing. degree from the University of Stuttgart, in 1962 and 1965, respectively. There he became Professor in 1972. Since 1977 he has been Head of the Laboratory of Control Engineering and Process Automation at the Institute of Automatic Control at the Darmstadt University of Technology, Germany; He acted as the chairman of several IFAC Symposia and also of the international program committee of the 10th

References (41)

  • Burg, J. P. (1968). A new analysis technique for time series data. In NATO Advanced Study Institute on Signal...
  • J. Chen et al.

    Robust model-based fault diagnosis for dynamic systems

    (1999)
  • R.N. Clark

    A simplified instrument detection scheme

    IEEE Transactions on Aerospace and Electronic Systems

    (1978)
  • Frank, P. M. (1987). Advanced fault-detection and isolation schemes using nonlinear and robust observers. In...
  • Freyermuth, B. (1993). Wissensbasierte Fehlerdiagnose am Beispiel eines Industrieroboters (Fortschr.-Ber. VDI Reihe 8...
  • R.A. Frost

    Introduction to knowledge base systems

    (1986)
  • Füssel, D. (2003). Fault diagnosis with tree-structured neuro-fuzzy systems (Fortschr.-Ber. VDI Reihe 8 No. 957)....
  • D. Füssel et al.

    Hierarchical motor diagnosis utilizing structural knowledge and a self-learning neuro-fuzzy-scheme

    IEEE Transactions on Industrial Electronics

    (2000)
  • J. Gertler

    Fault-detection and diagnosis in engineering systems

    (1998)
  • J. Gertler et al.

    Model-based diagnosis for automotive engines – algorithm development and testing on a production vehicle.

    IEEE Transactions on Control Systems Technologydkjdot

    (1995)
  • Cited by (0)

    Rolf Isermann received both the Dipl.-Ing degree in mechanical engineering and the Dr.-Ing. degree from the University of Stuttgart, in 1962 and 1965, respectively. There he became Professor in 1972. Since 1977 he has been Head of the Laboratory of Control Engineering and Process Automation at the Institute of Automatic Control at the Darmstadt University of Technology, Germany; He acted as the chairman of several IFAC Symposia and also of the international program committee of the 10th IFAC-Worldcongress on Automatic Control in Munich 1987, the IFAC-Symposium “Safeprocess” in Baden-Baden in 1991 and the IFAC-Symposium on “Mechatronic Systems” in Darmstadt 2000. In 1989, he was awarded by the Dr. h.c. degree from ĹUniversité Libre de Bruxelles and in 1996 from the University of Bucarest. In 2003 he received the Top Ten Award of emerging technologies from the MIT Technology Review Magazine for the field of mechatronic systems; R. Isermanńs main area of research interest covers: Process Modeling, Process Identification, Digital Control Systems, Adaptive Control Systems, Fault Diagnosis, Automotive Control, and Mechatronic Systems. He also wrote several books and papers on these topics.

    An earlier version of this paper was presented at the 16th IFAC Symposium on Automatic Control in Aerospace, St. Petersburg, Russia, June 14–18, 2004.

    View full text