Study On Parameter Identification Method Of 500kV Power Network Partition Load Model

With the development of DC feed,new energy grid connection and load development in recent years,the short circuit current,voltage stability,and dynamic stability of the system have become increasingly prominent.As an important factor affecting the above problems,the load is still used in simulation systems.Early typical model parameters,which may cause bias and economic waste in grid design,planning,and operation.Studying the load model parameter identification method has important practical significance for grasping the dynamic characteristics of the load,practicing the ubiquitous electric power network,and ensuring the safe and stable operation of the system.The popularization and development of wide-area measurement systems provide an important data basis for improving the simulation accuracy of load models.Based on the above background,this article mainly studies the following aspects:(1)Aiming at the problems existing in the parameter identification of 500 kV power grid zonal load model,the theory of zonal power supply of 500 kV power grid is studied,the advantages of overall measurement and identification method in parameter identification of load model are introduced,and the comprehensive load model structure considering distribution network branches Based on the summary analysis model,the sensitivity of the load model parameters to the positive sequence voltage,active power and reactive power of the system is analyzed by using the perturbation method and the similarity principle.(2)Aiming at the load model parameter modeling problem based on PMU(phasor measurement unit,PMU)measured data,the characteristics of PMU data are analyzed,and the load model parameter identification based on improved moth-flame optimization(MFO)algorithm is proposed method.This method uses the PMU master data for load modeling,and uses MATLAB and PSD-BPA joint simulation to complete the injection of the measured voltage data and the acquisition of active and reactive power simulation data.The goal is to minimize the sum of the squared errors between the measured and simulated values The function uses the MFO algorithm that introduces logistic chaotic map initialization,Lévy flight,and flame nonlinear quenching measures to design a parameter identification method for the load model.An example of a single load node demonstrates the effectiveness of the method.After comparison of multiple algorithms,the improved MFO algorithm overcomes the original algorithm’s initial value sensitivity,large moth randomness,and slow early convergence speed.The overall convergence and stability are relatively low.Strong.(3)Aiming at the current location of PMU devices,a method for parameter identification of 500 kV power grid zonal load models is proposed.From the perspective of 500 kV grid division,this method treats the load in the area as a whole,and uses the existing 500 kV substation and power plant PMU disturbance data.The power equivalent of the power plant based on power constraints and the external system voltage equivalent based on variable impedance To simplify the scale of the system,the design of the parameter identification method of the partition load model was completed,and a visual interface based on MATLAB-GUI was established.An actual grid ZZ partition example verified the effectiveness of the method.The influence of different load model parameters on short-circuit current,transient voltage characteristics,and dynamic stability characteristics confirms the importance of the load model to the safe and stable operation of the power system.