Remaining Useful Life Prediction for Spacecraft Actuator Based on Multiplicative Fault Observer

The actuator of attitude control system (ACS) plays an important role in providing the necessary control torque for spacecraft missions, such as orbital maneuvering and high-precision attitude stabilization. In a complex space environment and under long-term working loads, the actuator is more likely to degrade and fail compared to other components. This makes it significant to assess its reliability state and predict its remaining useful life (RUL). To overcome the drawback that traditional RUL prediction methods for the actuator are based on indirect external data, this article introduces a multiplicative fault from the control field as an intrinsic performance indicator to estimate the RUL. The multiplicative fault indicates the inherent degradation characteristic of actuation effectiveness, and its estimates can be obtained by designing a fault observer. Accordingly, a multiplicative fault observer is designed in this article, and its estimates are used to achieve the parameter estimation in the proposed generalized degradation model through maximum likelihood estimation (MLE). Furthermore, the probability density function (PDF) of the RUL is obtained. Illustrative examples of the multiplicative fault estimation and the corresponding RUL prediction are presented to demonstrate the effectiveness of the proposed method.