Fault Period Identification of Repairable Systems Based on the Model of Dual Weibull Distribution

The fault process of most repairable systems includes two fault periods. Hence, this article put forward a fault rate modeling and fault period identification method based on the model of dual Weibull distribution. Based on the same type test for multi-sample fault data, it is assumed that the fault rate follows the dual Weibull Distribution model. Then, the model parameters were estimated by combining the graphical method with the minimum sum of squared cumulative errors (T-L method). The model fitting was verified by K-S test, the optimal was determined according to root mean square error (RMSE), and the trend was analyzed using the U test to identify fault period. Finally, combined with the fault information of 7 domestic CNC lathes for two years, the case application proves the validity and rationality of the method.