The Study Of Centrifugal Compressor Fault Diagnosis Based On The Ant Colonies Algorithm

Stable and efficient operation of mechanical equipment to ensure production safety and production is an important prerequisite for quality, dynamic monitoring and fault diagnosis of equipment has a very high practical significance and analysis value, accurate trend forecasting, fault diagnosis and maintenance decision can not only reduce equipment maintenance costs, but also the risk of equipment accidents. Centrifugal compressor, which is as an important energy conversion device, has been widely used in the industrial sector, its reliability in working condition and operational has an important impact for the entire production system.As a rotating machine, the working condition of centrifugal compressor has a greater impact on the stable operation of the compressor. There are many reasons lead to a centrifugal compressor failure, processing errors, corrosion, mechanical damage of material properties, excitation, flutter, surge, wear, etc. are all the main influence of the centrifugal compressor work. This paper mainly study in three typical centrifugal compressor fault, oil whirl, surge and rotor imbalance, through analysis of radial compressor vibration signals under fault conditions, to study its vibration characteristics. And as a basis for fault diagnosis, further study of ant colony algorithm in centrifugal compressor fault diagnosis.Ant colony algorithm is a heuristic bionic algorithm, with a strong pattern recognition capabilities, that can analyze clustering data. In this paper, the ant colony algorithm model is established by MATLAB and carry out simulation experiments to characterize the measured vibration data. Through the fault diagnosis of rolling bearings to verify the feasibility of ant colony algorithm in fault diagnosis of rotating machinery, and to study its diagnostic features, and to optimize diagnostic methods further, analyze and diagnose the actual fault centrifugal compressor working conditions, compare the diagnostic results with the actual situation, the diagnosis is effective, stable and accurate.