Determination of Machine Condition Using Neural Networks

Book One

Volume

Proceedings of the AAAI Conference on Artificial Intelligence, 12

Student Abstracts

Condition monitoring is a developing discipline in machinery mainace. Data such as vibration levels, temperatures, oil analysis values etc, are acquired from plant, and analyzed to determine the condition of the plant at the time of measurement. Software packages are currently available to allow graphical display of the data, with varying levels of diagnostic tools available to assist engineers in performing data analysis. This abstract outlines the development of a condition monitoring system at Blyth Power Station, owned by National Power, the major electricity generating company in the United Kingdom. The abstract goes on to describe research into the development of a data analysis system employing neural networks trained to recognise machinery defects.