Efficient Photovoltaic Unit for Power Delivering to Stand-Alone Direct Current Buildings Using Artificial Intelligence Approach Based MPP Tracker

Abstract

There are many remote buildings that cannot be supplied by alternating electricity of the utility grid. Due to this, this study proposes adopting Direct Current (DC) appliances for a stand-alone remote building. Direct Current can be supplied from a suitable photovoltaic array which can harvest renewable solar energy. This proposal guarantees an efficient power system by removing the necessity of including an inverter, power filter, insulation transformer, and a complicated controller, which are usually needed for producing Alternating Current (AC) power to feed AC loads using a PV system. When the proposal is applied, the PV system will be more efficient, simple, affordable, and more compact. A detailed power requirement calculation for a typical house uses DC appliances, generalized steps to design a suitable PV array, and an Artificial Neural Network (ANN) algorithm for guaranteeing Maximum Power Point Tracking (MPPT); all of which are introduced for remote buildings. The main contribution of this paper is proposing an integrated design of a DC unit of 11 kW·h PV system for stand-alone buildings that eliminates three stages that improves the system performance compared to AC unit. The introduced study includes PV array calculation based on PV module of 220 W with an intelligent algorithm of four layers. The Mean Squared Error (MSE) of the proposed ANN equals 2.7107 × 10−5 to guarantee a fast and accurate MPP tracking for continuously harvesting maximum power from the incident sunlight. An energy storage unit of 12 batteries 12 V/150 Ah of matrix dimensions 3 × 4 is designed in the DC unit for energy saving to feed the DC appliances during night hours. MATLAB/Simulink Version R2015b is used to simulate the introduced DC power unit and collect the testing records for analyzing the unit performance.