Estimation-based approach to simultaneous adaptive control and parameter estimation in filtered-LMS problems

摘要

This paper presents an estimation-based approach to the adaptive filtering (control) problem in the filtered-LMS scenarios with unknown parameters in the secondary path. The approach discussed produces a robust estimate for the unknown system parameters while fulfilling adaptive control objectives. The control objective is to limit the worst-case energy gain from exogenous disturbances, initial condition uncertainty, and parameter estimation error to the cancellation error at the output of the secondary path. Simulation results on a typical active noise cancellation problem are presented to demonstrate the main features of the proposed approach.