Methods of grinding power signal acquisition and dynamic power monitoring database establishment

The grinding power monitoring experimental platform was built with PPC−3 power sensor and NI 9203 acquisition card. An intelligent grinding process decision-making system driven by monitored power data was developed based on LabVIEW software to promote green, efficient and intelligent grinding. In order to overcome the problems of huge amount of bottom process monitoring data (i.e. grinding dynamic power signals collected online) of the decision-making system, mixture with noise and unclear typical characteristics, a method of feature extraction of grinding power signals and establishment of relational database is proposed. The type Ⅱ Chebyshev low-pass filter was used to filter and improve the signal-to-noise ratio of grinding power signals. The peak and the valley characteristic points of power signals were extracted and marked in time domain based on the peak and the valley searching method, and the head and the tail correction and interpolation correction were carried out to ensure the integrity and accuracy of grinding power data. At the same time, the working state of grinding process was marked based on binarization, and the dynamic flow data was converted into string and stored in the cells of relational database. The grinding test results of bearing steel show that the database establishment method can accurately extract the grinding power characteristics and transform 2090000 dynamic data points into 2×52998 cell data, the data volume is reduced to 5.07% of the source data, which significantly reduces the storage scale of data and speeds up the access speed of grinding database.