| Microgird combines Distributed Generator(DG)and loads organically,which can effectively solve the problem of large-scale access of DG and improve the reliability of power supply.To advoid unintentional islanding,or to realize seamless switching between grid-connected and island modes,microgrid has a high requirement on the rapidity,reliability and micro-disturbance of islanding detection.The types of DG and the characteristics of load in microgrid will affect the dynamic change process of microgrid after islanding occurrence and are directly related to islanding detection method also.Microgrid with inverter-based DG,synchronous distributed generator and hybrid DG are studied respectively,and according to the different characterstic loads of microgrid,this paper investigates the active and passive islanding detection theories and a series of detection methods.The main contents of this paper are summarized as follows:1)The rapid and accurate detection of electrical signal parameters of microgrid ae the basis of islanding detection.Aiming at the problem that it is difficult to accurately estimate the frequency,amplitude and other parameters of the electric signal,the double suboptimal-scaling factor strong tracking Kalman filter(DSTKF)algorithm is proposed for estimating the electrical signal parameters of microgrid.First,taken into account the time-varying characteristics of frequency,amplitude of electrical signal of microgrid as well as impacts of dc,harmonics,a state space model of dynamic electrical signal is established by taylor series expansion.Secondly,Kalman filter algorithm fa ils to track the signal mutation when iteratively predicting the state variables at the next moment.The idea of strong tracking filter is introduced,and the mismatch degree between the theoretical and the actual residual in noise environment can be adjusted in real time by the fading factor s,which enhances the ability of DSTKF algorithm to respond to the abrupt signal.Simulation and experimental tests show that DSTKF has good dynamic response performance and measurement accuracy for signal mutation caus ed by islanding or fault occurrence.2)Aiming at the averaging effect of reactive power disturbance methods which can not guarantee the synchronization of the reactive power variations between inverter-based Distributed Generators(IDG),the active islanding detection method based on piecewise reactive power rule is proposed for microgrid.First,based on the analysis of the principle of reactive power disturbance,the disturbance and the system frequency deviation curve and the disturbance and load resona nce frequency curve,at the same time,by using the relationship between the system frequency deviation and reactive power disturbance after islanding occurrence,the piecewise reactive disturbance rule is constructed to eliminate the averaging effect.Secondly,the new islanding recognition criterion is constructed based on the frequency periodic fluctuation characteristic.The simulation tests show that the proposed method can identify islanding phenomenon without forcing the point of common coupling(PCC)frequency to exceed the specified operating range of the detection standard and avoid the disadvantage of IDG asynchronous reactive disturbance injection,and can be applied to the microgrid wit h the same control strategy IDG.3)The islanding mechanism of synchronous distributed generator(SDG)is seldom analyzed.Moreover,the existence islanding detection methods lack the basis for selecting the relevant characteristic s,which makes it difficult to select the reliable threshold.Therefore,the islanding detection method based on load characteristics is proposed for microgrid.Based on the analysis of equivalent circuit of SDG system,the difference of output impedance of SDG output terminal looking towards grid side before and after island ing is derived.At the same time,according to load characteristic,the islanding detection method based on the output impedance change rate is proposed,and the threshold value is set.Comparative analysis shows that the proposed method is better than other islanding detection methods,such as rate of change of frequency(ROCOF),rate of change of positive and negative sequence components of current and so on,and its non-detection zone(NDZ)is smaller.Even when the motor is started or the dynamic load is connected,t he proposed method still has strong reliability.4)Nonlinear load connected to microgrid with multiple IDGs,which makes the interconnection of microgrid characteristic and the islanding features a nonlinear variation,causing the existence passive method s to happen miscalculation.In addition,active methods can not be applied to microgrid with different control strategies IDG.For solving those problems,the passive method based on power mismatching boundary is presented.Based on the reconstruction of N DZ with dynamic and static characteristics,the power mismatch boundary characteristics before and after islanding difference is analyzed in detail for IDG system.At the same time,combining with the load characteristic,the specific expression of power mismatch boundary is derived,and the relative error of NDZ area is used as the islanding identification parameter to realize the rapid and accurate identification of islanding phenomenon,which overcomes the limitation of active method s.The simulation of short circuit fault,load switching,motor starting and islanding event verify the reliability of the proposed method.Moreover,the comparative analysis shows that the proposed method is better than some other island ing detection methods.5)For dealing with the problem that existence methods fail to identify islanding for microgrid with multiple IDG and SDG,the new passive method with adaptively threshold is proposed.Based on the analysis of the equivalent circuit of DG system,the time-domain expression of the derivative of the power factor angle at the PCC is derived.Then,according to the load characteristics,the island ing identification parameter is constructed.At the same time,the adaptive threshold is designed by analyzing the relationship betw een the reactive power mismatch and frequency offset.Finally,the applicability of proposed method to microgrid with multiple IDG and SDG is analyzed,which demonstrates the reliability of proposed method.The digital simulation tests verify the effectiveness of proposed method,and the problem of passive threshold selection is solved. |
PDF Full Text Request