An improved adaptive detection method for power quality improvement

A new adaptive harmonic selective detection method is proposed in this paper. This adaptive method has better convergence properties compared to previous adaptive or neuron detection methods. The adaptive gains of this method can be relatively large to obtain faster convergence. The stability of the proposed method is guaranteed and its properties are analyzed when compared to its predecessors. This proposed adaptive method is simple and effective in extracting fundamental and harmonic current information from nonlinear load currents. The extracted fundamental or harmonic currents therefore can be used as the reference signals for power quality improvement, harmonic selective cancellation or reactive power compensation. The method is also phase-independent and can thus be easily applied to single-or three-phase unbalanced systems. Simulation and experimental results verify the good performance of this improved adaptive detection method. Moreover, two application cases are built and tested to demonstrate its effectiveness in harmonic selective cancellation and reactive power compensation.