Modeling experiences of traditional grey-Markov show that the prediction results are not accurate when analyzed data are rare and fluctuated.So it is necessary to revise or improve the original modeling procedure of the grey-Markov(GM)model.Therefore,a new idea is brought forward that the Markov theory is used twice,where the first time is to extend the original data and the second to calculate and estimate the residual errors.Then by comparing the original data sequence from a fault prediction case with the simulation sequence produced by the use of GM(1,1) and the new GM method,results are conforming to the original data.Finally,an assumption of GM model is put forward as the future work. Modeling experiences of traditional gray-Markov show that the prediction results are not accurate when analyzed data are rare and fluctuated. So it is necessary to revise or improve the original modeling procedure of the gray-Markov (GM) model.Therefore, a new idea is brought forward that the Markov theory is used twice, where the first time is to extend the original data and the second to calculate and estimate the residual errors.Then by comparing the original data sequence from a fault prediction case with the simulation sequence produced by the use of GM (1,1) and the new GM method, results are conforming to the original data. Finally, an assumption of GM model is put forward as the future work.