Health Aware Control Design for Extending Remaining Useful Life of Electro-Hydrostatic Actuator Based Steering

Electro-hydrostatic actuators (EHAs) are used in aerospace, industrial, and off-road applications due to their efficiency and self-contained operation.
However, degradation of critical components, such as the pump, can lead to reduced performance and unexpected failures.

Conventional control strategies often neglect real-time degradation effects, limiting their ability to optimize both actuator performance and remaining useful life (RUL).
This study addresses this gap by developing a health-aware control (HAC) framework that integrates degradation modeling with predictive control for EHAs.

The proposed HAC system incorporates a nonlinear Wiener degradation model to estimate internal leakage progression in the pump.
The degradation state is estimated online using an extended Kalman filter (EKF) and integrated into a model predictive control (MPC) framework to dynamically adjust control actions based on degradation predictions.
The effectiveness of the approach is evaluated through simulations, comparing it to a baseline PID-controller.
The choice of MPC and EKF over traditional PID control is justified by their ability to handle nonlinearities, provide predictive capabilities, and offer robust state estimation.

Simulation results show that the HAC system extends the RUL of the gear pump by \qty{3.22}{\percent}, compared to the PID controller.
The system successfully mitigates degradation effects, while maintaining an average position tracking error of \qty{1.39}{\milli\meter}.

Although the RUL improvement is modest under accelerated degradation conditions, the proposed approach shows strong potential for real-world applications where degradation occurs over extended periods.

Future work will focus on experimental validation using physical test rigs and refining the degradation model to enhance prediction accuracy and control