Research Of Micro-grid Islanding Detection Method Based On Hilbert Huang Transform Algorithm

After loss of the traditional centralized power supply, distributed powergeneration system and the local load run independently, which are calledislanding.Islanding may give rise to secondary reclosing, or even damage the micro-grid system, load and power supply system. Thus, to ensure power quality, itdemands timely detection of island status.Then distributed generation system isdisconnected from the point of common coupling and optimized combination ofdistributed generation systems and supply patterns continue to supply power tothe load.Micro-grid islanding detection is one necessary function of grid-connectedinverter system. Islanding detection process is premise of cutting off or re-optimization of distributed generation system. Pros and cons of detectionperformance is directly related to the safe operation of equipment, and thepopularity of micro-grid technology. Time-frequency analysis method is used inislanding detection topics.This paper proposes a passive islanding detectionmethod based on Hilbert-Huang transform algorithm by summarizing the existingislanding detection method, according to the collection and identification fromvoltage signals micro-components at the point of common coupling. Microscopiccomponent is high frequency characteristics components of the voltage signal, onthe time of micro-grid switch mode. The new detection method is mainly basedon unequal power quality between inverter side and main grid side. During theswitch mode, harmonic of signal is more obvious.To extract characteristic frequency signal when islanding occurs, thoughEmpirical Mode Decomposition and Hilbert transform, so as to determine thetime and location of singularities and islanding. To build photovoltaic distributedgeneration systems by PSCAD/EMTDC and MATLAB/SIMULINK software,so that it can realize smooth handoff between parallel operation and islanding operation. Through final simulation results, demonstrate the validity andaccuracy of the HHT algorithm in islanding detection topics. The simulationresults also shows the method can quickly detect islanding under various types ofvoltage fluctuation and load operation, switching, even other disturbances.