Under the new energy development goal of "peak carbon dioxide emissions and carbon neutrality",our country’s "14th Five-Year Plan" clearly proposed to promote the energy revolution,build a clean,low-carbon,safe and efficient energy system,which will further accelerate the power electronics of our country’s power system.In order to ensure the safe and stable operation of the power system containing a large number of power electronic equipment,electromagnetic transient simulation is required.Power electronic equipment has the characteristics of complex control and operating characteristics,high operating frequency,and fast transient process,so that when the power system containing a large number of power electronic equipment is simulated,it is necessary to carry out refined small-step simulation of power electronic devices,but also to simulate other devices with longer dynamic processes in the power grid,which poses new challenges and requirements for electromagnetic transient simulation.Therefore,it is necessary to study the fast electromagnetic transient simulation method of the large power grid.This paper aims at this and the following research work has been carried out:(1)The selection and design of the numerical integration algorithm affects the performance of electromagnetic transient simulation,and its characteristics need to be analyzed before the selection and design of the numerical integration algorithm.Therefore,this paper studies the representative simulation algorithms and analyzes their characteristics in simulation accuracy,stability,applicability and numerical oscillation.At the same time,At the same time,the characteristics of these simulation algorithms are summarized,and on this basis,combined with the characteristics of electromagnetic transient simulation,the basic requirements and implementation directions for selecting and designing simulation algorithms are obtained,which lays a theoretical foundation for the subsequent proposal of the discrete similarity exponential integral method.(2)Starting from the discretization nature of simulation,a discrete similarity exponential integration method is proposed.This method constructs discrete system similar to continuous system in the frequency domain using pole-zero response matching,which is difference with the traditional simulation that the differentiation is directly made in the time domain.And then inverse transformation of discrete system in the frequency domain to obtain difference equation,starting from the causes of numerical oscillations,combined components capable of effectively suppressing numerical oscillation is obtained through analysis and comparison.For independent inductor and capacitor that cannot be directly differentiated by using pole-zero response matching,they are solved by series and parallel positive and negative virtual resistors respectively,thereby ensuring the method can be used as a global simulation method.In addition,the relevant characteristics of the discrete similarity exponential integral method are also studied to ensure the effectiveness of the method in simulation.(3)From the point of view of taking into account both simulation accuracy and speed,an adaptive variable step size simulation strategy is proposed,and an adaptive multi-band dynamic phasor model is given by combining the adaptive variable step size strategy with the multi-band dynamic phasor model.The model simulates the converter based on the discrete similarity exponential integration method,which can take into account various working conditions of the converter,using large step simulation in steady state conditions and small step size simulation in transient conditions.Finally,the adaptive multi-band dynamic phasor model and its simulation flow are given,and the characteristics of the model are summarized and analyzed. |