Digital Twin-Driven Condition Monitoring for Servo Valves

To address the challenge of real-time monitoring of internal parameters such as spool pressure in servo valves for aircraft flight control actuators, which is primarily due to the compact structural design, a state monitoring method driven by a digital twin (DT) framework is proposed. High-precision Computational Fluid Dynamics simulation data are utilized to construct a Radial Basis Function surrogate model that establishes a real-time mapping between measurable external parameters, including drive voltage, spool motion cycle, and temperature, and the internal pressure field. The model's accuracy is validated by calculating the coefficient of determination (R2). Combined with the K-Nearest Neighbors algorithm for spatial feature mapping, a DT of the servo valve is developed using the Unity3D engine. The results show the established DT precisely predicts the high-fidelity simulation outputs of the maximum pressure, with the maximum relative error remaining below 0.50 percent. Furthermore, the discrepancies in the position predictions of the maximum pressure between the simulation and the DT are generally within 5.00 percent. This implementation enables visual tracking and dynamic prediction of flow field evolution within the valve, thereby achieving comprehensive internal flow monitoring and seamless interaction between virtual and physical systems.