Islanding Detection Technology Research On Grid-connected Photovoltaic Power Generation System

With the rapid development of photovoltaic(PV) industry and national policies are in support of PV connecting with grid, PV connected-grid technology has become one of the key research area. To achieve safe and reliable operation, grid-connected PV system must have islanding detection ability. The research is mainly on islanding detection technology with single-phase grid-connected PV system platform in this paper.Firstly, this paper analyzes the mathematical model of islanding detection and IEEE standard. It combines Low Voltage Ride Through(LVRT) requirements and islanding operation time together to give a more reasonable operation time for grid-connected PV system for the first time. Non-detection zones(NDZ) theory has been systematically analyzed.Secondly, although passive islanding detection technology has blind spot detection, it is applied widely as it is simpler and more affordable. The voltage amplitude and frequency extraction is a key point for passive islanding detection technology. This paper focuses on theoretical analysis and simulation studies on three voltage amplitude and frequency extraction methods which includes DPLL(Digital Phase-Locked Loop), STF(Strong Tracking Filter) algorithm and SOGI-FLL(Second Order Generalized Integrator-Frequency Locked Loop), meanwhile, gives comparative analysis for its strengths and problems. Considering the fast algorithm, accuracy and stability, it applies multi-channel SOGI-FLL for passive islanding detection for the first time and builds a single- phase photovoltaic islanding detection simulation model which demonstrates multi-channel SOGI-FLL can extract the amplitude and frequency accurately and quickly, thereby it can effectively detect the occurrence of the island.What is more, passive islanding has detection blind spot detection problem under the power matching, active islanding detection intends to break the power balance by adding a disturbance and then detects the islanding. First, it focus on theoretical analysis and simulation research of traditional AFD(Active Frequency Drift) method. The simulation results proves its strengths and problems. Based on the above research, this paper proposes a new islanding detection method that is intermittent frequency perturbation method, the algorithm uses a multi-channel phase-locked loop section SOGI-FLL which can extract voltage amplitude and frequency more quickly and accurately; it gives frequency disturbances in fixed circles avoiding the zero-crossing detection disturbance problems; intermittent disturbance improves harmonic distortion due to continuous disturbance, it also adds the amplitude disturbance, frequency and amplitude perturbations work together to further improve the reliability of islanding detection, experiments prove its effectiveness of the algorithm through simulation experiments.Finally, this paper presents a design based on programmable DC source to simulate high power solar array and builds an experimental prototype. Through the closed-loop and open-loop system comparison experiments. It proves the solar array simulators to be effective and feasible. While building a small single-phase grid-connected PV system hardware test system, and complete the grid connecting experiment. It provides a hardware platform for further study of islanding detection.