Experimental evaluation of an active power line conditioner based on a neural network control scheme

The purpose of this paper is to evaluate a proposed voltage-type active power line conditioner (APLC) consisting of a neural network controller, a variable DC voltage source, and a DC/AC inverter. The APLC is a type of active filter which is a very efficient apparatus for compensating harmonics caused by nonlinear loads. Computer simulations of the proposed APLC have been performed using MATLAB. The results show that the proposed APLC can reduce the current total harmonic distortion (THD) of a specific nonlinear load from 54% to less than 5% and improve the power factor close to one. A low power rating prototype model has been built with commercially available components to implement the proposed APLC. Measured results clearly support the simulation results and suggest that the proposed system is a simple and effective approach to line current harmonic compensation