Research On Modeling Method Of Frequency Response In Multi-power Hybrid System

Frequency is one of the important indexes of power system operation,which reflects the energy balance relationship between power system and load.Frequency and its dynamic characteristics play an important role in the normal operation of equipment and the safe operation of the power grid,as well as stable control.Therefore,this paper studies the frequency response characteristics and modeling methods of the multi-power hybrid system.This work mainly includes four parts,the first part uses the modular approach to model the thermal power unit.Modular modeling divides the thermal power unit into several basic modules such as coordinated control system(CCS),speed control system(DEH),steam turbine,and boiler according to the system function.They are modeled and simulated in Simulink separately.We can obtain the thermal power unit simulation model by combining the modules above.Based on the simulation model,the dynamic characteristics of the primary frequency response of the thermal power unit are studied.The second part combines the equivalent model with the radial basis function neural network(RBFNNs)to establish the aggregate model of the wind-fire hybrid power system.The equivalent model is used to determine the structure of the RBFNNs,which lays a foundation for simplifying data-driven modeling basis.The third part develops a wind-thermal hybrid power system identification algorithm based on Discreet State Space Neural Network(DSSNNs).Considering the existence of non-Gaussian disturbances in the system,the survival information potential(SIP)is selected as the performance index of the identification algorithm,the simulation results show that the frequency response model of the wind-thermal hybrid power system established by the discrete state space neural network is reasonable and effective,and analyze the frequency response characteristics of the hybrid power system under different wind transmittance.The fourth part focusing on the problem of lacking data sample in wind-thermal hybrid power system.The identification method based on transfer learning(TL)and recurrent neural network(RNNs)is proposed to establish the frequency response model of the system.Using the wind speed and load to describe the scene,and the frequency response of the wind-heat hybrid power system operating in different scenes is analyzed through simulation experiments.