A model reference adaptive control based method for actuator delay estimation in real-time testing

Abstract

Real-time testing provides a viable experimental technique to evaluate the performance of structural systems subjected to dynamic loading. Servo-hydraulic actuators are often utilized to apply calculated displacements from an integration algorithm to the experimental structures in a real-time manner. The compensation of actuator delay is therefore critical to achieve stable and reliable experimental results. The advances in compensation methods based on adaptive control theory enable researchers to accommodate variable actuator delay and achieve good actuator control for real-time tests. However, these adaptive methods all require time duration for actuator delay adaptation. Experiments show that a good actuator delay estimate can help optimize the performance of the adaptive compensation methods. The rate of adaptation also requires that a good actuator delay estimate be acquired especially for the tests where the peak structural response might occur at the beginning of the tests. This paper presents a model reference adaptive control based method to identify the parameter of a simplified discrete model for servo-hydraulic dynamics and the resulting compensation method. Simulations are conducted using both numerical analysis and experimental results to evaluate the effectiveness of the proposed estimation method.

Keywords

real-time testing / actuator delay / compensation / adaptive control / MIT rule / discrete transfer function