Categories

A sample text widget

Etiam pulvinar consectetur dolor sed malesuada. Ut convallis euismod dolor nec pretium. Nunc ut tristique massa.

Nam sodales mi vitae dolor ullamcorper et vulputate enim accumsan. Morbi orci magna, tincidunt vitae molestie nec, molestie at mi. Nulla nulla lorem, suscipit in posuere in, interdum non magna.

What are the different fault detection techniques?

Fault detection and isolation is categorized under the field of control engineering, which is concerned with the monitoring of a system and identification of a fault wherever it has occurred. It also pinpoints the location of the fault or error. Fault detection is also supposed to specify the type of the fault. Basically for fault detection, two approaches are followed:
1. Direct pattern recognition of sensor readings: These readings indicate a fault and an analysis of the fault between the readings of the sensor and the expected values which have been derived from some model.
2. A fault is said to be detected if the discovered residual or discrepancy goes above a certain limit. It is then categorized according to its type by the techniques of fault isolation and its location is also specified in the machinery or the software system.
Fault detection and isolation (FDI) techniques are broadly classified into two categories:
1. Model-based FDI: In this technique, some models of the software system or hardware system is used to decide the type, location and occurrence of the fault. Some models may be mathematical and some may be based on knowledge. Some models also include parity space approach, observer based approach and parameter identification based approach.
2. Signal processing based FDI: In this technique some mathematical operations are performed using measurements to get the knowledge of the fault.
There has been a rising demand for better fault detection techniques as a result of growing demand for better and higher performance, and also due to increase in the level of automation and also efficiency, reliability and safety in both hardware and software systems. Diagnosis proves to be a complex reasoning activity and is currently one of the fields where Artificial Intelligence methodologies have been successfully implemented. These techniques make use of association, decision making processes and reasoning as human beings would in solving these problems.
Classical fault detection methods are entirely based up on checking of the limit of the values of some important measurable variables. These methods require a lot of research work. One of the drawbacks of the classical fault detection methods is that these methods do not permit for an in-depth fault diagnosis and do not simulate the activity of human reasoning. Some powerful new techniques have been developed that have the capability of reasoning about deep models, and to operate with a great deal of information. Artificial Intelligence experiments conducted with the help of models of human intelligence and by making systems that exist autonomously in their assigned environment are able to act intelligently and efficiently. Applications such as the following are being used as fault detection techniques:
1. Expert systems
2. Neural networks and
3. Intelligent signal processing
One of the Current trends involve coupling of the above mentioned applications in order to produce more effective fault detection tools. For the purpose of diagnosis, the knowledge-driven techniques are used, and they involve the interpretation of signals from the sensor, detection of abnormal activities, and generation of hypotheses about the behavior and nature of the fault and also provide fault explanation. Mathematical modeling techniques that come under classical numerical fault detection techniques have the advantage of allowing fault prediction and detection of faults in an early phase of development. Fault detection has both economical and safety benefits as it helps in the prevention of future program failures and improves the schedules of program maintenance.
Sometimes mathematical modeling techniques can be combined with artificial intelligence techniques in order to improve the effective coupling of symbolic and numerical information. These days quantitative analytical models of the algorithms and analytical methods of fault detection are being combined with the qualitative models of knowledge-based techniques.

Fault Detection: Theory, Methods and Systems Software Fault Tolerance Techniques and Implementation Model-based Fault Diagnosis Techniques: Design Schemes, Algorithms, and Tools

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>