Research On Fault Diagnosis Of Intermittent Reactor Temperature Control System

Condition monitoring and fault diagnosis technology includes two steps, they are monitoring and diagnosis. In modern industry, the large inertia time-varying delay systems' fault diagnosis is so difficult, because that it is far away from monitoring the fault event precisely. So the paper gives a new algorithm, named grouping trend of derivation algorithms, to monitor the equipments'state.Paper gives brief details of this new algorithm and the steps of realization of fault diagnosis. First step, the algorithm will group the system data in accordance with the law. Second step, analyze the character of every grouping of data. Third step, orientate the fault using some fault diagnosis methods. This algorithm has several features. Firstly, because the system sampling time is stationary, the fault diagnosis system can catch the signal of the fault event and get the accurate time. Secondly, fault diagnosis which using trend prediction based on grouped data is more accurate than using the system model which generated by all sampled data. Thirdly, this algorithm only need to work with one group data every time, so the hardware load is much lower. At last, for a different object system, the algorithm only need to modify some parameters to realize the fault diagnosis, and this algorithm is much more universal.Article conduct research on the large inertia time-varying delay systems' condition monitoring and fault diagnosis which using the grouping trend of derivation algorithm. And the result of the experiment is quite satisfactory and effective.