Power-Flow-Based Stabilization for Adaptive Feedforward Filters in Hybrid Testing

Abstract

Real-time hybrid testing is a technology which allows the coupling of simulations and component tests in order to simulate complex system dynamics. Delays and time lags caused by actuator dynamics and signal processing deteriorate the stability of the tests in many cases. The application of adaptive feedforward filters to hybrid testing enables circumventing this problem. The stability of the filter itself, however, can be affected by the choice of the algorithm parameters or changes in dynamics of the system being tested. Test safety requirements and practical considerations require a failsafe implementation. In this paper, we propose a method for adjusting the parameters of the adaptive feedforward filter based on power-flows in the test setup. The objective is to maintain a passive behavior of the actuation and control system. The stabilization acts on the leakage factor and the adaptation gain of a least-mean-squares adaptation law. A simple numerical system is used to investigate the effect of the algorithm parameters on the stabilization. The method was applied to an experimental setup including a nonlinear stiffness. Several originally unstable configurations were stabilized, the adaptation process could be continued and interface synchronization was achieved in all test cases.