Multi-information fusion diagnosis of lubrication oil contamination using fuzzy distance

摘要

The lubrication oil used in equipments is needed to be tested and the oil contaminated state is needed to be diagnosed and employed to make a decision whether to change the oil. The test methods have different effects on the oil attributes and the different data process methods also result in different test results. To overcome the limits, the diagnosis methods are improved to consider these factors and employed to get more precise result, and give more appropriate suggestion. A fuzzy method are used to solve the above problems, and the test conditions, the tested attributes and the data process method are regarded as evidences to judge the contamination state by decision makers, rather than the numeral computation method is only employed to calculate the attributes level using single test values, and the contamination state degree is simplified and idealized. Further, the information imperfectness and the decision makers weights are also solved by a new decision making approach using fuzzy distance and evidence theory proposed for the multi-attribute group decision making problem with the fuzzy support assessment information given by different decision makers. The weights of decision makers are established and adjusted by the fuzzy distance, which are employed to fuse the support assessment information integrated by D-S theory of evidence. An example of the lubricant diagnosis is used to illustrate the validity and the application of the method.