Study Of Micrology Shape Analysis And Automatic Identification Technique For Wear Particles

As one of the most effective methods for machine condition monitoring and fault diagnosis, ferrography technique has already played a great role in industrial application. However, with the fast development of modern industry, the traditional ferrography could not satisfy the real demand due to its subjectivity, low precision and time consuming, even limited its wider application. Therefore, the development of automatic ferrography is essential for overcoming the current technique obstacles. The automatic ferrography has also been a focus research topic in recent years.Debris automatic identification and analysis system (DAIAS) based on Visual C++6.0 software platform by studying image processing technique of computer and pattern recognition method and tribology knowledge has been developed in the present research. In this system, wear particles were signed in image by pretreatment of color debris image firstly. Secondly, the quantification character parameter system of debris micrography was produced by studying recognition character of debris. At the same time, three types of character parameters (color, surface texture and shape & size) were confirmed. Then, a parameter database was founded with extracted numerical descriptors of debris. Finally, the theory of fixed-weight-grey-clustering was used to classify six types of wear particles (rubbing, spherical, cutting, severe sliding, Fe2O3 and Fe3O4 wear particles) successfully according to their morphology numerical parameters analysis.The applied image examples showed the methods in the present thesis were briefly and effectively, which provided an important basis for the further study of machine fault diagnosis.