Online Assessment Of Power System Transient Stability Based On Random Matrix Spectrum

Power system transient stability assessment has always been one of the important research contents to ensure the safe and stable operation of power system.With the continuous expansion of the power grid and the access of a large number of renewable energy sources,the dynamic behavior and operation modes of the power system are getting complex,and there are stricter requirements for transient stability assessment.The wide application of Wide Area Measurement System(WAMS)provides abundant data resources for transient stability online assessment based on WAMS data.In this paper,the big data technology and the transient stability mechanism are combined,and the transient stability analysis index based on the measurement data is proposed to realize the on-line analysis and assessment of the transient stability of the power system.WAMS data is the basis for data-driven based on-line analysis of power system transient stability.According to the transient response characteristics and physical mechanism of power system,this paper extracts characteristic information from WAMS data that can fully reflect the transient stability response process.Moreover,the construction method of characteristic information data model based on random matrix theory is proposed.In order to realize the real-time monitoring of transient stability and the assessment of disturbance severity in the transient process.This paper comprehensively uses the node characteristic information and generator characteristic information to establish a random matrix model for transient stability analysis,and formulates the spectral deviation index for transient stability analysis based on the spectral distribution characteristics of ring law and M-P theorem.The disturbance severity is quantitatively assessed according to the integral of the index over the fault duration.On this basis,the comprehensive spectral deviation index is constructed,and the transient stability of the power system is monitored and analyzed online by combining discrimination mechanism of system stability of random matrix theory and the variation law of the comprehensive spectral deviation index in the disturbance process.For multi-machine systems,the augmented matrix method is used to formulate the degree of interference evaluation index from the perspective of data correlation,to quantitatively evaluate the degree of disturbance of different generators when the system fails.According to the evaluation results,the identification strategy of critical generator pairs is formulated to realize rapid identification of critical generator pairs in the system when faults occur.Based on the above research,aiming at the influence of redundant historical data on spectral distribution characteristics in the engineering application,this paper uses the insertion matrix method to improve the real-time separation window,and takes the generator pair relative kinetic energy as the influencing factor to build the data model of power system transient response characteristics.The maximum eigenvalue deviation(MED)difference trajectory representing data correlation is obtained.Then the mapping relationship between Med difference trajectory and transient power angle stability under different stability modes is summarized and the transient power angle stability criterion is formulated.An online transient stability assessment method of multi-machine power system based on the critical generator pair spectral distribution characteristics is proposed.This method does not need to cluster the generators in the analysis process,simplifies the transient stability discrimination process,makes the assessment faster,and has certain anti-noise ability.It can be applied to stability assessment under complex disturbances.