Modelling and Statical Analysis of Fixed and Tracking Photovoltaic modules in the Highveld of South Africa

This paper presents the approximate normal and descriptive statistics results for the fixed-types and dual-axis sun tracking-types systems located in the Highveld of South Africa. The fixed PV modules are installed at the orientation angle of 0° and tilt angles of 16°, 26° and 36° as recommended by Heywood and Chinnery respectively. The dual-axis tracking systems make use of two algorithms (fuzzy logic and linear regression) for automatic movement. To fulfill the specifications needed for a dual-axis tracker, the algorithms are developed in a National Instruments (NI) LabVIEW user interface software to control three linear actuators. The knowledge base of a fuzzy logic algorithm consists of a database and a rule base. The fuzzy control rules are generated in order to track the maximum intensity of voltage that is directly related to the amount of direct beam radiation from the sun. Linear regression algorithm is used to determine the optimal tilt and orientation angles for a PV module by modelling the relationship between two variables (scalar-Y and denoted X). The two variables represented by voltage (Y-axis) and time (X-axis) are used to control the movement of the PV module for optimum efficiency. The PV modules rotate automatically based on the output power during the day while at night; the system is in sleep mode in order to reduce the energy consumption. A computer model and data collection for the systems are modelled using LabVIEW user interface software. Results suggested normal distribution with kurtosis and skewness being within ±1.