Wear Condition Monitoring System Based On Oil Sample Spectral Analysis Equipment

For a long time, People are always paying attention to the problem of monitoring equipment status in manufacturing in order to know about the equipment operation status and oil quality. And they made a lot of research, such as vibration, yawp, temperature, oil, sound emission and lossless detection technology, and so on. Among of these, oil analysis technology is an important one, which is fit for analysis of equipment status in liquid or semi-liquid condition.In the background of the corporative project between ZHUJIANG STEEL L.T.D and GUANGZHOU MACHINE SIENCE RESEARCH INSTITUTE, the paper researches on the spectral analysis technology of oil. And the result is applied to judging equipment acute wear and wear place. The paper introduces the development status and base analysis technology of oil and anatomizes the difficulties. Then, for this type of problem, a relatively integrated solution is put forward on the basis of assembling the relative methods. And it better solves this kind of practical application problems.Choosing the 1# hydraulic LP system of casting and rolling in ZHUJIANG STEEL L.T.D as the researching object, the paper provides the method of appropriately dealing with initial data and judging the acute wear of equipment by using PLS arithmetic which is fit for the condition of fewer samples and more variables. Based on principle of equipment wear, an appropriate explaining for PLS's analysis result is made. If the equipment is in condition of acute wear, the variable group will be determined by using Fuzzy Clustering Arithmetic according to relativities of all elements under different level, and the pattern of waiting for identifying will be constructed. As concerned as multiple elements relativity, the concept of variable group and variable group formicate are put forward. Of course, the paper gives the Fuzzy Cluster Arithmetic principle, process and clustering chart. Then, possible wear place is judged by using Gray Association Degree theory plus the above variable group. In one word, for the practical application problems, the paper gives the methodology solutions to this type of problem on the basis of assembling different methods. The result shows that theAbstractmethod is right and useable to some extent.Also, the paper gives the system analysis and design, realized method, some functional examples and introduction. In the end, summary of the paper and a vista of future are provided.