Artificial neural networks for harmonic currents identification in active power filtering schemes

This paper presents a new harmonic currents identification method called neural synchronous method and based on artificial neural networks. Its theoretical aspect relies on a new decomposition of the load current signals. Adaline neural networks are used in order to learn this decomposition on-line; the fundamental currents can therefore be estimated at each sampling time. The fundamental currents are then synchronized with the direct component of the voltage obtained by a PLL (phase locked loop). The harmonic currents are deduced and re-injected phase-opposite in the power distribution system through an active power filtering scheme. This harmonic currents identification method is compared to other similar methods by simulation results.