Real-time investigation of the effects of digital filter methods on power analysis in a grid connected PV system

Reference current and voltage values, especially the control structures of inverter and converter circuits, must be measured with high accuracy to improve the power quality in grid-connected PV systems. The high accuracy of the power analysis in the control structures of the converter circuits will result in the error-free calculation of other electrical parameters such as active power, reactive power, and phase angle of the PV system, thus increasing the power quality of the system. The power analysis block is generally used in the literature to accurately measure the voltage and current signals at the PV inverter output and to obtain other electrical parameters. In addition, although analog filters are a practical solution to reduce harmonics on the grid side, they bring additional losses to the power losses caused by inverters and converters, and a hardware change can only be made in the system. In the design of digital filters, the changes to be made in the filter can be made by easily changing the parameters in the software, not the hardware as in analog filters. For this reason, there is a need to use software-based digital filters in this area, especially in PV systems, to improve power quality by performing power analysis and calculation with high accuracy. In this thesis, digital filters are used to contribute to the power calculation by minimizing the distortions in the electrical signals, instead of the existing power analysis block in the literature, to perform power analysis in PV systems with high accuracy. The effects of different types of digital filters, performed both in simulation and in real-time, on power analysis and calculation in the LabVIEW environment were investigated, the digital filter type with the highest accuracy was determined, and this developed structure was integrated into a grid-connected PV system. In the simulation part of the thesis, a pure 50 Hz sinus signal representing the PV system inverter output in the LabVIEW environment is distorted by white noise, and 8th-order bandpass Butterworth, Bessel, Elliptic IIR, and LMS algorithm digital filters are applied to the harmonic signal, respectively. Although all filters were successful in noise removal in the simulation, Butterworth and Bessel's filters were more successful in noise removal than other filters. In the last stage of the thesis, the proposed approach was also carried out experimentally in the Renewable Energy and Smart Grids Laboratory at Bursa Technical University. The real-time current-voltage signals acquired from the grid-connected inverter output with a total power of 1200 Wp connected to the solar panels were transferred to the data acquisition card, and the effects of digital filters on the power analysis were investigated in real-time with the software implemented in the LabVIEW environment. 8th order bandpass Butterworth, Bessel, and Elliptic IIR digital filters were developed as software in the LabVIEW environment and applied to the current and voltage signals of the inverter output of the grid-connected PV system in the laboratory, current, voltage, active power, reactive power, apparent power and power factor in the system. The measured values in real-time were also verified with the power analyzer and it was seen that the Butterworth and Bessel numerical filters gave results very close to the real measurements. With the method proposed in the thesis, instead of the traditional calculations in the power analysis block in a grid-connected PV system, a real-time approach with high accuracy and enhanced with digital filters is carried out. Thanks to this method, which enables power analysis with high accuracy by preventing noise in current, voltage, phase difference and power signals in power analysis, control structures in inverter and converter circuits of grid-connected PV systems operate much more stably. Thus, a structure that contributes to the improvement of power quality by supporting the suppression of harmonics by analog filters has been introduced. The structure realized in the thesis study applies not only to the power system but also to power electronics and motor drivers, hysteresis control, d-q conversion, etc. In terms of increasing the stability of many control structures, it offers a structure that can be developed for different areas in practice.