or a cylinder gap or leakage (multiple faults) of 1st
class compressor stuffing box. For the data of 3rd
group, the diagnosis result is the cooler water system
fault. The result of the experiments is consistent with
the actual fault and indicates that the fuzzy relation
matrix obtained is reasonable and the diagnosis result
is effective.
4 CONCLUSION
(1)The fault diagnosis system based on .NET platform
achieves good application effect by the test of simu-
lation, field data and fault analysis and realizes fault
diagnosis located in wellhead and compressor.
(2) The fuzzy neural network associative memory
model (FNNAM) for the compressor fault diagnosis
has achieved good effect during the self-learning of
fuzzy relation matrix. The FNNAM algorithm depends ...