A remaining useful life prediction method was proposed based on Kalman filter and expectation-maximization algorithm to solve the estimation problem of remaining useful life of helicopter's main retarder with introduction of indirect condition information. The proposed method can estimate the model parameters rapidly and effectively by updating characteristic value of vibration signal

and then remaining useful life distribution of main retarder was estimated at different working times. Finally

the case analysis was conducted through the test data of main retarder. The results demonstrate that the proposed method can effectively estimate the remaining useful life distribution of main retarder. Through comparison of real remaining useful life and estimated remaining useful life of main retarder

it is found that the estimated remaining useful life is mainly contained within 95% confidence interval

showing the proposed method has well accuracy

which can avoid the occurrence of the fault.