Research On Islanding Detection Method Of Grid-connected Photovoltaic Power Generation System

Since the 21st century, due to the highlighted traditional energy crisis problem, the development and utilization of the new energy has become the focus of attention. The grid-connected photovoltaic (PV) power generation is a form of effective utilization for the new energy, and has been got rapid development and wide application. The islanding effect detection is one important issue of the grid-connected techniques which connect the PV power generation system to grid must be solved. If the islanding effect is not eliminated in time, it will endanger the safety of personnel, equipment and the stability of system. Therefore, research to achieve an effective islanding method has been a popular project in recent years.According to the research background of this topic, first, the mechanism of islanding occurrence, international detection standards, effectiveness evaluation and classification of islanding detection methods are hackled and summarized. Then the concepts, the advantages and disadvantages for various types of islanding detection methods are emphatically analyzed and summarized, which set the stage for later in-depth study work for new detection method.Then, two works have been completed in the paper through deeply analyzing the active phase shift (APS) islanding detection method and the wavelet transform islanding detection method which belongs to passive method.(1) Because of the issues of larger non-detection zone (NDZ) and greater influence on the output power quality existed in the classical APS islanding detection method, an improved APS islanding detection method is proposed based on the fuzzy control theory in the paper. The feedback coefficient k of the improved method is adjusted adaptively by the detection method, compared with the classical method, not only the detection speed has been improved, the NDZ has been reduced, but also the distortion of inverter output current has been reduced by 0.48%, that is the influence of disturbances to power quality is reduced. Meanwhile, the relationship between the parameters and the NDZ of the method has been analyzed in detail in the paper. The simulation results base on Matlab/simulink verify the effectiveness and superior performance of the improved method.(2) There are some shortages when the real wavelet transform and wavelet packet transform which are used to detect islanding, such as tedious layered feature and complex process to determine the threshold of feature. In order to solve these issues, an islanding detection method based on complex Morlet wavelet transform is proposed in the paper. The islanding detection has implemented via extracting and analyzing the changes of the amplitude and phase of complex wavelet coefficient for phase-voltage at the point of common coupling (PCC) by the proposed method. The proposed method has advantages that are faster detection, simple to determine the threshold and non harmonic interference. The simulation results based on Matlab/simulink have verified the effectiveness and superior performance of the proposed method in case of the harmonic interference state, load(s) sudden change state and electric power system fault case.In conclusion, the islanding effect detection can be effectively implemented by the research of the active phase shift islanding detection based on the fuzzy control and the passive islanding detection method based on complex Morlet wavelet transform in the paper. And the detection performance of the islanding has improved in two methods. The research has some certain theoretical significance and application value in the islanding detection.