Investigations On Fault Detection Methods And Protection Princlples For Microgrids

The application of distributed generator(DG)in microgrids results in some new problems in modern power industry.Solar power outputs are influenced by natural climate,and hence are random,intermittent,and volatile.It may cause voltage fluctuation in the distribution system concerned.The extensive applications of photovoltaic inverters may cause harmonic pollution.For a large number of distributed grid-connected photovoltaic,the distribution system will become an active network.Thus the power flow and the voltage distribution in the system would change.Especially the rise of voltage is beyond the limit.With the distributed generators,especially renewable energy,constantly improving the proportion of traditional power system,microgrids are highly be focused on worldwide,which is regarded as a new energy integration by the demand side.And it is an effective way to improve the efficiency of comprehensive utilization of energy.More specifically the protection challenges arise due to the high penetration of DG with power electronics interfaces in the microgrid.The inertia of inverter based DG units is lower as compared to synchronous machines.A microgrid consisting of low inertia sources might have stability issues in case line faults are not cleared fast.Typically,the maximum current contribution capacity of IBDGs is around two times of per unit rated current.So,the fault current level for the microgrid operating in islanded mode of operation is considerably smaller as compared to the grid-connected mode for any type of fault as the IBDGs have limited current carrying capacity.Fault protection becomes challenging using conventional high fault current approach.The power flow in a microgrid is dynamic and bidirectional as the microgrid has DG and also exchanges power with the utility grid.Protection problem gets complicated when the microgrid switches between mesh and radial topologies.Specifically,the contents of this dissertation are as follows:1)Employing the correlative influence factors with the detection and protection for the microgrids.Firstly the technology standards and regulations are surveyed;Then the topology structures,the grounding modes and the control schemes with DG are reviewed.Next,the influence from the different penetrance rates and locations of DGs are analyzed.The definition of a 'Unit-based microgrid topology' is put forward.On this basis,the transient characteristics when different faults occurred are analyzed.Finally the applicability of conventional protection principles is discussed.2)Investigating the fault feature selection approach which is applicative for both islanded and grid-connected operation mode.The algorithm of machine learning is employed.Given a series of fault feature based on conventional electric parameters,Support Vector Machine(SVM)and k-Nearest Neighbor(k-NN)are combined and implemented in this framework.To get the optimal minimum feature subset,this feature selection method reduces the feature dimension,decreases the system complexity,and makes storage space smaller.At the same time this method has good accuracy and generalization ability.This may afford the basis for fault detection and protection schemes for the microgrid subsequently.The proposed approach is tested on the benchmark microgrid model.3)A novel method to detect the interior fault in a microgid with Hilbert-Huang transform.First a typical microgrid mode is built up,which can switch between islanded operation and grid-connected operation.And different faults are set.Measure the voltage at the monitoring point.Then the Hilbert-Huang transform is applied to analysis by MATLAB.Finally the monitoring point's waveform of voltage,the corresponding intrinsic mode function diagram,the instantaneous amplitude,instantaneous frequency,amplitude spectrum and marginal spectrum may all be obtained.Compare the fault condition with the normal fault condition,and analysis their differences.This detecting approach is proved to be valid.Finally by the contrast experiment it is verified that Hilbert-Huang transform is more adaptive than Wavelet transform for microgrid fault detection.4)A new protection principle based on the negative sequence components of both the bus voltage and the impedance.There are some advantages of them.For instance,they are not related to the connection mode of the micro sources and the isolated transformer.Also they are not influenced by the load impedance.Therefore it is good to utilize the negative sequence components of some electrical parameters for the sensitivity and the reliability of a microgrid.Through an analytical approach,the changing law of the impedance angle of negative sequence component is deduced,which may contribut to identifying the fault in the microgrid.This method is proved to be effective by the test experiments.Finally,some conclusions are obtained based on the research outcomes,and directions for future research indicated.