Prerequisite/Recommended prerequisite for
participation in the module
Probability, statistics and stochastic processes, system
identification and estimation
Content, progress and pedagogy of the
module
Purpose
- To contribute to students’ attainment of comprehension of some
typical fault detection and diagnosis techniques.
Content
- Fundamental concepts, terms and principles of FDD
- Fault modelling and analysis
- Fault types and classification
- Fault modelling
- Fault delectability
- Fault diagnosability
- Residual generation (I): Observer based FDD methods for
deterministic systems
- Review of observer theory
- Fault detection using single observer
- Fault diagnosis using a bank of observers
- Residual generation (II): Kalman filter based FDD methods for
stochastic systems
- Review of probability and stochastic processes
- Kalman filter theory
- Extended Kalman filter
- Fault detection using single Kalman filter
- Fault diagnosis using a bank of Kalman filters (Multiple Model
(MM) method)
- Fault diagnosis using a bank interactive Kalman filters
(Interactive Multiple Model (IMM) method)
- Fault diagnosis using a two-stage Kalman filter for additive
and multiplicative faults
- Robust residual generation (I): Unknown Input Observer (UIO)
method
- (complete) Disturbance decoupling principle
- UIO theory
- Robust FDD using UIO method
- Robust residual generation (II): Robust filtering method
- Disturbance attenuation principle
- Modelling uncertainties
- Introduction to robust filtering theory (H_infty optimal
control theory)
- Robust FDD using H_infty filtering method
- Residual evaluation
- Simple voting techniques
- Statistical testing approaches
- Likelihood function methods
- Probabilities of false alarm and miss
- FDD using Parity space approaches
- Delectability and diagnosability
- Parity space methods for FDD
- Parameter estimation based FDD methods
- Parametric fault characteristics
- FDD using parameter estimation (least-square methods)
- FDD using recursive system identification
methods
- Signal-based (model-free) FDD methods
- FDD using spectrum analysis
- FDD using short-timed Fourier transform and wavelet
transform
- FDD using some artificial intelligence
methods
Learning objectives
Knowledge
- Have comprehension of some typical model-free fault detection
and diagnosis methods
- Have comprehension of some typical model-based fault detection
and diagnosis methods
Skills
- Are able to apply the learned knowledge to handle some fault
detection and diagnosis problems.
- Are able to judge the usefulness of the set up methods
- Are able to relate the methods to applications in the
industry
Competences
- Independently be able to define and analyze scientific problems
within the area of fault detection and diagnosis.
- Independently be able to be a part of professional and
interdisciplinary development work within the area of fault
detection and diagnosis.
Type of instruction
The program is based on a combination of academic,
problem-oriented and interdisciplinary approaches and organized
based on the following work and evaluation methods that combine
skills and reflection:
- Lectures
- Classroom instruction
- Project work
- Workshops
- Exercises (individually and in groups)
- Teacher feedback
- Reflection
- Portfolio work
Extent and expected workload
Since it is a 5 ECTS course module, the work load is expected to
be 150 hours for the student
Exam
Exams
Name of exam | Fault Detection and Diagnosis Techniques |
Type of exam | Written or oral exam |
ECTS | 5 |
Assessment | 7-point grading scale |
Type of grading | Internal examination |
Criteria of assessment | The criteria of assessment are stated in the Examination
Policies and Procedures |