Research on oil monitoring data modelling and prediction based on grey theory

Accurate modeling and prediction of oil debris monitoring data for equipment power systems plays an important role in evaluation of the health status of power systems and ensuring their normal operation. An improved two-step GM (1,1) modeling method for non equidistant sampling of oil monitoring parameters, which is often encountered in engineering practice, is studied. This method is not only suitable for high growth sequence, but also has high model accuracy. Based on this method, the grey prediction model of oil wear particle number is established and the prediction accuracy is compared with that of the non-equidistance modeling accuracy of data transformation method. The results show that the former has higher accuracy and is more suitable for the short-term prediction of oil wear particle number trend.