An adaptive prediction error filter for photovoltaic power harvesting applications

Abstract

This paper proposes a new adaptive prediction error filter (APEF) for harvesting maximum power from a photovoltaic (PV) system. The device used for maximum power harvesting application is known as maximum power point tracker (MPPT). Power harvesting efficiency is directly dependent on this MPPT; hence there is a necessity of a properly designed MPPT. An adaptive MPPT is capable of efficient operation for a wider operating range of the PV system such as solar irradiances and temperature. An adaptive MPPT usually required two components i.e.; a system identifier of the PV system and a controller. The proposed APEF-MPPT can perform both the role system identifier and controller. It also considers external disturbances at the designing stage, hence operates efficiently in the presence of disturbances. The APEF-MPPT consists of a linear predictor that tuned on-line with environmental conditions. The efficacy of the proposed APEF-MPPT is investigated using a PV test system and found that it provides accurate and robust performance.