Comparison of Parameters Estimation Techniques Applied to an FCS-MPC Controlled HAPF

As any other model dependent control technique, the finite control set model predictive control (FCS-MPC) cannot properly control the system when a large difference between the real and the considered parameters values occur. Taking that into account, this article performs a comparison between three techniques to estimate the most sensitive parameter on a hybrid active power filter (HAPF) applied for reactive power compensation. The first technique was recently proposed and is based on the application of adaptive notch filters (ANF) to the system state variables, the second technique is the recursive least squares (RLS) algorithm and the third is an adaptive RLS algorithm. The comparison considers the steady-state error, over-shoot, settling time and computational burden of each technique. Both simulation and practical results are used in the analysis.