Research On Digital Oil Monitoring Based On Ferrography Technology

Wear can lead to consumption of material which affect the service life and reliability of mechanical equipment. Ferrography technology which regards wear particle as the research object is one of important technologies in the state monitoring of mechanical equipment. This paper is aimed at problems that different conclusions may be drawn when people look at the same ferrogram, digital research of wear particle image is studied by classifying and counting, Specific studies are as follows:(1) A reasonable and effective method for the segmentation of wear particles was found. RGB color feature segmentation method based on region growing which can remove the background was determined by researching on different segmentation methods, this method reduces the error of artificial points; the growth rules were determined by the regional growth experiment of wear particle images; Ostu segmentation method was used to separate the grain from the sedimentary chain.(2) Characteristic parameters of wear particles were analysed. The characteristic parameters (size, color, texture) of wear particles image were extracted, the influence degree of different parameters on the recognition of wear particles and the parameter range were determined by analysing parameters.(3) An effective algorithm for the identification of the type and composition of wear particles was proposed. The recognition theory (BP neural theory, grey correlation analysis) was put forward by analyzing the principle, advantages and disadvantages of common intelligent recognition theory.(4) The recognition software was written by using Visual Basic 2008 to solve the problems which were mentioned, and basic features were realized. The segmentation results and statistical results by using the actual wear particle images showed that the segmentation algorithm and the recognition method were reasonable and effective.