Research On Transient Stability Of Microgrid Based On Virtual Inertial Control

Under the condition that large-scale of renewable sources are integrated to the grid,virtual inertia control method represented by droop and virtual synchronous generator is effective for improvement of inverter’s transient stability and the research frontier in inverter control method.However,the inverter-dominated power systems are facing with transient angle stability problem due to virtual inertia and the nature is different from that of traditional power grid.This problem has attracted the attention of some experts and related research has been carried out.But the current study is still inadequate in understanding the unique characteristics of the inverter-dominated system and the transient behavior of grid-connected inverters,nor the quantitative stability margin index.Even if there are relevant qualitative analysis methods,the error will be large due to inadequate analysis.This paper is focused on the issue of transient angle stability of microgrid that is based on virtual inertia control.Firstly,the physical mechanism and mathematical essence of grid-connected inverter stability are discussed by comparing the inverter and generator system.Based on that,two key factors are concluded as virtual damping and saturation mechanism.Secondly,grid-connected inverter’s transient behavior of each control loop is analyzed in order of time-scale and a complete mathematical expression for transient stability analysis is given.Then a quantitative analysis method of transient stability for gird-connected inverter is proposed by referring to EEAC and considering dissipative term and saturation mechanism.What’s more,a stability margin index is given and the influence of control parameters such as droop coefficient and virtual inertia coefficient is clarified.Thirdly,in order to extend single-machine system to multi-machine system,a data-driven assessment algorithm for transient stability of multi-inverters power system is proposed with the application of deep learning theory.Through the design of comparative tests,the selection of system features is optimized and the accuracy of stability prediction is improved.Finally,the two transient stability assessment strategies are simulated on the platform of MATLAB or Keras.And simulation results verified the effectiveness.There are three points of innovation work in this paper:1)The transient behavior of grid-connected inverter is analyzed in detail and the analytical expressions are derivated in order of time scale.And the differential equations are summarized for transient stability analysis.2)A quantitative stability analysis method is proposed to make the stability analysis more accurate by combining the advantages of time domain simulation and direct method.3)A fast and high accuracy stability assessment strategy of multi-machine system is realized with the application of deep learning theory in this research topic.The significance of this work is summarized as follows.Firstly,it provides a transient stability analysis method for the operation of large-scale integration into grid of renewable energy in the future.Secondly,the analogy between the inverter and generator system and the transient behavior analysis at different time scales provide a theoretical basis for the capacity design and energy storage allocation of the inverters.Thirdly,the influence of the key parameters of the inverter on the system transient stability is clarified,which provides a basis for the inverter’s real-time parameters control.Fourthly,the combination of deep learning and power electronics provides a new research idea,which can be extended to other issues in the field of power electronics or power systems in the future.