Research On Parameter Identification Of Synchronous Generator Based On PMU Information

The main feature of modern power systems is the introduction of modern control theory and computer technology,but both control and calculation require accurate and reliable mathematical models.As a main component of the power grid,the synchronous generator is built on a precise model to facilitate the safe and stable operation of the system.The traditional parameter identification method is gradually eliminated due to the inability to make into account the actual operating conditions of the system.The online identification based on Synchronous Phasor Measurement Unit(PMU)has become the main research method in the current identification field.Although researchers have made relevant research on the parameter identification of synchronous generators based on PMU data in recent years,they have not studied the model parameter identifiability analysis and model accuracy evaluation based on PMU data.In this paper,the parameter identification of synchronous generator based on PMU data is carried out,and the parameters identifiability analysis algorithm,model parameter identification and model accuracy evalution are systematically studied.The specific research contents are as follows:1)The problem of identifiability of synchronous generator parameters based on PMU data is studied.A identifiability analysis algorithm for synchronous generator parameters based on Principal Hessian Direction(PHD)theory is proposed.Firstly,the target function of synchronous generator parameter identification is established,and the feasibility analysis of parameter identification is performed based on the objective function;Then the PHD algorithm is used to propose the parameter identifiability analysis scheme to avoid the local saddle point and the parameter unrecognizable problem.Finally,the correlation parameter is analyzed and identified.2)Study on online identification of synchronous generator parameters based on PMU data,Firstly,the data information is formatted,the data length is selected and preprocessed.Then the d and q axes are decoupled and identified for identifiability analysis.Finally,the effectiveness of the method is verfied by several actual case data.3)Research on the model and parameter accuracy evaluation based on PMU data,and propose a method based on multi-scene perturbation data to estimate the accuracy of the comprehensive similarity index.Firstly,the external motor is considered to be equivalent to the external network,and the power signal is injected for hybrid dynamic simulation.Then,the comprehensive similarity index is calculated for multiple single disturbance fault data,and the accuracy of the model is judged based on the threshold value.Finally,the error is large.Make further parameter corrections.Simulation cases show the effectiveness of the method.