Advantage analysis of PI-based voltage control for grid-forming converter transitioning via phase portrait analysis

With increasing renewable energy sources, they are often faced with operating modes that require them to transition between islanding and grid-connected modes at any time. However, since the active power and reactive power modulation loop in grid-forming converter have the coupling effect for each other to provide the reference voltage and frequency, this can lead to instability phenomenon in the transition process. Thus, the simple PI-based voltage direct control scheme can be taken as a candidate in a scenario that requires frequent transitions. Its transient stability is analyzed by design-oriented to prove its advantage in this situation in this paper. Since the nonlinearity of its transfer function, a Deep Reinforcement Learning method is selected to accomplish the PI parameter tuning task. The test results show that the simple PI-based voltage control approach along with its optimal parameters has ability to deal with the transition process to keep stability in low SCR grid conditions. • PI-based voltage control has advantages in certain scenario. • Phase portrait is validated for transient stability analysis. • Deep-learning algorithm is used for tuning parameters.