The oil atomic spectrometric analysis technology is one of the most important methods for fault diagnosis and state monitoring of large machine equipment. The gray method is preponderant in the trend forecast at the same time. With the use of oil atomic spectrometric analysis result and combining the gray forecast theory, the present paper established a gray forecast model of the Fe/Cu concentration trend in the power-shift steering transmission. Aiming at the shortage of the gray method used in the trend forecast, the improved Euler algorithm was put forward for the first time to resolve the problem of the gray model and avoid the non-precision that the old gray model’s forecast value depends on the first test value. This new method can make the forecast value more precision as shown in the example. Combined with the threshold value of the oil atomic spectrometric analysis, the new method was applied on the Fe/Cu concentration forecast and the premonition of fault information was obtained. So we can take steps to prevent the fault and this algorithm can be popularized to the state monitoring in the industry.