Objective
Objectives and content
This research aims to furnish a methodology and the corresponding computer code for a diagnostic system which, when coupled to existing monitoring facilities for rotating machinery in power plants, would allow the nature, location and severity of faults and malfunctions to be identified with a high degree of confidence. The research is timely because most operators of large power plants have invested in modern computer based vibration monitoring systems which collect and store large amounts of data. Current technology does not make use of all the quantitative information in this data.
The proposed diagnostic system, in contrast to all existing diagnostic tools, will be based on models of the rotor-bearing-foundation system and models of possible faults or malfunctions. Faults will be detected by comparing on-line data from plant and simulated output from the model using advanced computational methods. The system will provide the plant operator with better quantitative information concerning the plant condition, allowing more informed commercial and safety decisions regarding plant operation to be made. It would also allow residual life of plant to be estimated; an extremely important feature since within the European Community many power plants are more than 20 years old and approaching the end of their design life. The technology could be transferred to rotating machinery users in other European industrial sectors.
The research will consist of the following main activities:
- Definition of models:
Models will be prepared for components and for the complete
rotor-bearing-foundation system using system identification and parameter estimation techniques where required to maximise model accuracy. Models will include non-linearities where appropriate, they will incorporate sensitivity information to quantify their robustness and they will be optimised for efficient computation.
- Definition and models of faults:
Fault models, compatible with the system models, will be developed for each of the principal failure modes considered and used in the model based diagnostics.
- Experimental verification:
The reliability of the system and fault models will be verified using test rigs and full scale in-situ measurements on the industrial partners' power plant under various operating conditions.
- Development of model-based diagnostic methodology:
An intelligent diagnostic system, using advanced computation methods such as a neural network and/or knowledge based system will be developed. It will interface with all the available machine monitoring instrumentation together with data from the system and fault models. Part of this work will include an assessment of monitoring instrumentation and signal processing requirements.
Fields of science (EuroSciVoc)
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsignal processing
- natural sciencescomputer and information sciencescomputational science
- natural sciencescomputer and information sciencessoftwaresoftware applicationssimulation software
- natural sciencescomputer and information sciencesartificial intelligencecomputational intelligence
You need to log in or register to use this function
Call for proposal
Data not availableFunding Scheme
CSC - Cost-sharing contractsCoordinator
B4 7ET Birmingham
United Kingdom