Modelling of SFR for Wind-Thermal Power Systems via Improved RBF Neural Networks

Integrating Physical and Data-Driven System Frequency Response Modelling for Wind-PV-Thermal Power Systems

“…Support vector regression was employed to estimate the minimum frequency and dynamic SFR of a disturbed power system [16]. Based on survival information potential (SIP) criterion, an improved radial basis function neural network (RBFNN) was presented to build the equivalent model of windthermal integrated power systems [17]. Although nonlinearities and uncertainties of the SFR might be revealed using the above data-driven modelling methods, however, their accuracy depends on the quantity and quality of the database.…”

“…Neural networksbased system identification methods have been used to model complex processes or systems with nonlinearities and uncertainties and demonstrated their versatility and effectiveness [19][20]. In particular, the improved RBFNNsbased system identification approaches were presented and applied in [17,[21][22][23][24][25]. Along this line of consideration, an improved radial basis function neural network (RBFNN) is utilized to build a data-driven model in this work, which deals with non-Gaussian disturbances from PV and wind power plants.…”

SFR Modeling for Hybrid Power Systems Based on Deep Transfer Learning

“…Investigation on power system frequency response (SFR) can reveal the dynamic characteristics of system frequency under the disturbance of generation or load, some approaches to SFR modelling have been presented [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]. The existed power system frequency response modelling approaches can be classified as three categories: direct measurement method [3][4][5][6], mechanism analysis method [7][8][9][10][11][12][13][14][15] and data-driven method [16][17][18][19].…”

“…Authors in [18] utilized support vector regression (SVR) to predict power system frequency dynamics and its nadir after disturbances. In [19], an improved radial basis function neural networks was employed to model SFR of a wind-thermal hybrid power system (HPS). The data-driven model in [16][17][18][19] described frequency behavior in the vicinity of a certain scenario based on collected input-output data.…”

“…In [19], an improved radial basis function neural networks was employed to model SFR of a wind-thermal hybrid power system (HPS). The data-driven model in [16][17][18][19] described frequency behavior in the vicinity of a certain scenario based on collected input-output data. The accuracy of the data-driven model is related to the number of samples available to train the model.…”