A Neural Lyapunov Approach to Transient Stability Assessment in Interconnected Microgrids

We propose a neural Lyapunov approach to assessing transient stability in power electronic-interfaced microgrid interconnections. The problem of assessment is cast as one learning network-structured function the state space. Based on learned, security region estimated for monitoring interconnected microgrids real-time operation. efficacy tested and validated grid-connected three-microgrid interconnection. A comparison study suggests that proposed method can achieve less conservative characterization region, compared with conventional approach.