Research On VSG Control Strategy For Suppressing Low Frequency Oscillation Of Power System Based On Transient Energy Flow

Energy scarcity and environmental degradation have gradually become the object of global attention.The development of renewable energy power generation technology is the fundamental means to solve this problem.A large number of new energy grid-connected power generation equipment are connected to the power grid through power electronic converters.Due to the characteristics of power electronic converters such as fast response,weak damping,and parametric resonance,oscillation is very likely to occur in grid-connected power systems with high proportion of renewable energy,which seriously endangers the safety of the system.As a control strategy of grid connected inverter,virtual synchronous generator（VSG）can actively participate in oscillation suppression as a controlled device and improve the stability of power system operation.Considering the internal relationship between transient energy flow and power system stability,it can provide theoretical support for VSG to participate in system low-frequency oscillation（LFO）suppression.Therefore,this paper analyzes the VSG oscillation mechanism around the transient energy flow,and studies the VSG suppression strategy of power system low frequency oscillation.The main research contents are as follows:（1）The theoretical basis of LFO based on VSG transient energy flow suppression system is analyzed.Firstly,the calculation formula of VSG transient energy flow is deduced and its influence on system stability is analyzed.According to the calculation formula,two key variables affecting the transient energy flow of VSG are determined.The amplitude frequency characteristics and phase frequency characteristics between them are derived through small signal analysis,so as to clarify the influence of VSG control parameters J and K_d on the magnitude of transient energy flow.Finally,the correctness of the derivation is verified by a simulation example,and the idea of low-frequency oscillation suppression by adjusting J and K_d to control VSG to absorb transient energy flow as much as possible is established,which provides theoretical support for subsequent research.（2）Since the rapidity of LFO modal identification is of great significance in this study,a power system LFO modal feature identification method based on deep learning is proposed.In order to meet the requirements of LFO suppression strategy design for rapid identification of LFO modal features,a deep learning algorithm is introduced.In the aspect of network model,LSTM neural network is selected and sliding window sampling is used to realize the rapid identification of non-stationary LFO modal characteristics.In order to avoid the order-determining process of LFO,in terms of classification criteria,a binary classification method based on existence verification is adopted,which means the samples are classified according to the presence or absence of a certain modal feature.Simulation examples show that the proposed method is more effective than the traditional LFO modal analysis method in identifying the modal features and modal change process of various LFOs including non-stationary LFOs.（3）An LFO control strategy for VSG suppression system based on transient energy flow is proposed.Firstly,referring to the design idea and with the help of LFO mode fast identification algorithm,the control parameters J and K_d are determined with the goal of VSG absorbing transient energy flow as much as possible,so as to design a parameter adjustment strategy of VSG suppression system LFO.Then,simulation experiments are carried out in the four machine two area system and various high proportion renewable energy grid connected systems.Compared with the adaptive strategy,it can be proved that the proposed strategy is more flexible and effective.The good performance of the proposed strategy in each system proves that it has a good application prospect in the environment of high proportion of renewable energy grid connection.