Normal Harmonic Search Algorithm-Based MPPT for Solar PV System and Integrated With Grid Using Reduced Sensor Approach and PNKLMS Algorithm

This paper deals with a novel reduced sensor strategy, for two-stage single-phase grid connected solar photovoltaic system with a battery, where the dc-link voltage sensor is not required. Moreover, for maximum power point tracking (MPPT), a new “normal harmonic search” (NHS) algorithm is used, and for voltage source converter (VSC) control, a novel “power normalized kernel least mean square” (PNKLMS) algorithm is proposed. The PNKLMS algorithm is used for active load component extraction from the load current, even when the load current is characterized by linear or nonlinear behavior. Moreover, PNKLMS algorithm is used for reference grid current generation. The NHS algorithm is an improved version of the harmony search, which is inspired to the music composition technique and its searching ability is improved by using the normal probability distribution factor. Therefore, NHS reaches the global maximum power point very quickly and is suitable for online searching on hardware. Moreover, the extracted power is fed into the grid. For control of VSC, the proposed PNKLMS control technique is used, which maintains the power quality. The performances of NHS-based MPPT and PNKLMS-based control technique with reduced sensor approach are tested on a developed prototype. The satisfactory steady-state and dynamic performances of the NHS algorithm are demonstrated under variable irradiance and temperature during partial shaded conditions. Moreover, the performances of the PNKLMS algorithm, without using dc-link voltage sensor are demonstrated during overvoltage, undervoltage, harmonics distortion, and load nonlinearity condition, results of which have satisfied the objective of the proposed system and the IEEE-519 standard.