Modeling and Simulation of Electrically Excited Synchronous Machine Based on Discrete State Event-Driven Approach

Microgrids can realize the flexible access of distributed generations, energy storages, and loads. A typical microgrid system often contains various DC and AC power supplies and many power electronic devices, challenging conventional simulation tools. Recently, the discrete state event-driven (DSED) approach has been proposed and can efficiently solve large-scale power electronic converters. This method models power electronic converters as piecewise linear (PWL) state equations and consists of a flexible, adaptive integration method based on the Taylor series. It can solve PWL systems but cannot solve the system containing electrically excited synchronous machines. Therefore, this paper proposes a state-variable-interfaced decoupling strategy based on the DSED approach. The state-variable-interfaced decoupling strategy can decouple the electrically excited synchronous machine from the whole system. The electrically excited synchronous machine and the other part systems exchange high-order time derivatives of interface variables to assure the simulation accuracy. In addition, a method for simplifying the time derivative calculation of electrically excited synchronous machines’ state variables is proposed. The time derivatives can be obtained recursively, and the complicated and troublesome high-order derivative expressions can be avoided. The numerical errors introduced by the simplifying method are analyzed and can be well controlled using the proposed error equations. In the case study of a microgrid system, compared with commercial simulation software, the proposed method can improve the calculation speed by more than 40 times with the same accuracy. The simulated microgrid system consists of two distributed synchronous machines as AC generations and three DC generations with inverters. The simulated working conditions are as follows: initially, only DC power supplies are available, and the microgrid is connected to the power grid; At 0.2 s, put AC power into operation, that is, electrically excited synchronous machines; At 0.4 s, the microgrid is disconnected from the power grid, and the AC power supply supports the line voltage and frequency. The simulated results show good agreement with the commercial simulation software. The improvement of computing efficiency mainly benefits from two aspects. The single-step computational cost of the adaptive integration method is lower than that of the commonly used explicit Runge-Kutta method under the same order. The DSED method considers events and uses an event-driven strategy to conduct numerical integration of continuous state variables between event points, reducing the number of calculation points. The following conclusions can be drawn from the simulation analysis: (1) When adopting the ideal switched model, power electronic systems contain discrete switch events and continuous state events. The DSED method adopts a flexible, adaptive Taylor series-based integration method to adapt to this hybrid nature and achieve high simulation efficiency. (2) The proposed state-variable-interfaced decoupling strategy makes the DSED method efficiently solve the microgrid with electrically excited synchronous machines. (3) The proposed method has the potential to be transferred to solve systems containing other motors, such as induction motors. © 2023 Chinese Machine Press. All rights reserved.