Multi-sensor information fusion-based prediction of remaining useful life of nonlinear Wiener process
Bin Wu, Jianchao Zeng, Hui Shi, Xiaohong Zhang, Guannan Shi, Yankai Qin
Abstract
Abstract The use of multi-sensor information fusion techniques is essential for condition monitoring and prediction in large complex systems. In this paper, a new distributed model fusion method is proposed to predict the remaining useful life (RUL) of a nonlinear Wiener process. First, the state–space model of the nonlinear Wiener process is established, based on multi-sensor monitoring, and the distributed Kalman filtering algorithm is used to filter and fuse the measurement data received from multiple sensors. Next, the parameters and degradation states of the state–space model are estimated and updated online in real time using the expectation maximum and smoothing filter algorithms. Moreover, the distribution of the system’s RUL is obtained according to the estimated state–space model considering the random failure threshold factor. Finally, numerical experiments are conducted to elucidate the accuracy of the adopted distributed fusion method, and the adaptability and effectiveness of the proposed method are verified using the FD001 data of the C-MPASS dataset as an example.