Improved Transients in Multiple Frequencies Estimation via Dynamic Regressor Extension and Mixing

摘要

A problem of performance enhancement for multiple frequencies estimation isstudied. First, we consider a basic gradient-based estimation approach withglobal exponential convergence. Next, we apply dynamic regressor extension andmixing technique to improve transient performance of the basic approach andensure non-strict monotonicity of estimation errors. Simulation resultsillustrate benefits of the proposed solution.