Research On Low-frequency Oscillation Type Discrimination Method In Power Systems

The low frequency oscillation of power system is one of the most important problems that threaten the safe and stable operation of power network.It is of great significance to accurately determine the type of oscillation for reasonable suppression measures and improving the dynamic stability of the power grid.The key of the research on classification study of low frequency oscillation is the type classification and the effective criteria.This paper reviews the research results in the domestic and foreign related fields firstly and then introduces the mechanisms,analytical approaches and the control measures of low frequency oscillation.The existing oscillation type distinguish methods have also been mentioned in this paper.On this basis,the problems of oscillation type classification are deeply analyzed and the effective criteria are put forward in this paper.Firstly,from the macroscopic point of view,the oscillation type is subdivided into local oscillation,interval oscillation and local-interval coupled oscillation.The oscillation causes,suppression measures,basic analysis methods and performance characteristics of different oscillations has been analyzed.The classification of low frequency oscillation is studied from the point of view of energy,and the oscillation energy of low frequency oscillation is analyzed both on the spatial view and the time view.The space-time analysis method of oscillation energy is proposed to extract the different characteristics of different types of oscillation.The case of MengXi Power Grid verified the rationality of this method.The low frequency oscillation type can be discriminated by the different characteristics of oscillation energy in the space-time distribution.In order to quantitatively characterize the different characteristics of different types of low frequency oscillation,the entropy theory and the coefficient of variation theory are introduced to the research on the oscillation type discrimination.The criteria of spatial distribution entropy and normalized standard deviation are put forward and the low frequency oscillation types can be distinguished effectively by the entropy-deviation classification criterion.Finally,the classification and discrimination of local oscillation are studied.This paper divides the local oscillation into parametric-local oscillation and the network-local oscillation by analyzing the causes of negative damping.The discrimination method of local oscillation types is studied based on network information and the static stability margin criterion is put forward.Calculating the static stability limit by the time-domain simulation method is not suitable for on-line analysis.So,this paper study the estimation method of static stability limit based on the WAMS measurement information and Thevenin equivalent method.Grid example verified the validity of the local oscillation classification method and the accuracy of the static stability limit estimation method.