A method of sampling point optimization in fault diagnosis
Huiling Liu · 2017
To gain the fault information and determine the fault position, monitoring point should be selected properly, otherwise, it usually fails to fulfill the task effectively and accurately in monitoring and diagnosing the gearbox. A method based on condition attribute reduction technology in Rough Sets was proposed to optimize the sampling point. First, the decision table was established according to Rough Sets Theory. Further, an improved NaiveScaler algorithm was put forward to discrete decision table, a method based on the discernibility matrix and attribute frequency was presented to calculate the minimal attribute reduction sets. Finally, the most sensitive signal monitoring point was achieved through analyzing the final reduction sets. The results show that it is feasible to apply the attributes reduction technology to select the sensitive sampling points, and the attribute reduction technology needs neither modeling for the monitoring object nor dynamics analysis, but selects the effective sampling point directly according to the relationship between the time-frequency domain parameters and fault types, so it is also simple and convenient to optimize the measuring points.