This article presents an implementation of an adaptive learning based backpropagation (AL-BP) control technique for an improved power quality grid-interactive solar photovoltaic (PV)–battery supported system catering the balanced and unbalanced three-phase four-wire (3P-4W) nonlinear loads and single-phase loads of various nature simultaneously. Moreover, the source currents remain balanced and sinusoidal for all operating and loading conditions, such as in highly unbalanced loading conditions where a load on one of the phases is disconnected. An AL-BP control technique is used to control the voltage at point of common coupling and to improve power quality under various loading conditions through mitigation of harmonic currents, reactive power compensation, active power balancing, and neutral current compensation in the 3P-4W system. It also improves the system power factor in highly nonlinear and unbalanced loading conditions. The AL-BP control scheme has the capability to self-tune the arbitrary nonlinear systems, such as a grid-interactive critically unbalanced 3P-4W system feeding highly nonlinear loading systems. Therefore, the proposed control technique is found capable of dealing with the above-mentioned large unknown nonlinearity of the 3P-4W system. This scheme significantly improves the steady-state and dynamic performances of the 3P-4W grid-interactive solar PV–battery supported system. The proposed AL-BP control technique has been implemented using a DSP processor in real time in order to validate the claims.
Read full abstract