An Interruption Overvoltage Estimation Method With Improved Swarm Intelligence Algorithm for Modular DC Circuit Breakers

Jie Gao, Huan Yuan, Aijun Yang, Mingzhe Rong, Xiaohua Wang

Abstract: Modular dc circuit breakers (MDCCBs) are the key equipment for dc networks, and their interruption overvoltage is an important aspect of protecting and monitoring MDCCBs. This article analyzes the existing problems of the MDCCB interruption overvoltage estimation from two aspects, including a metal-oxide varistor (MOV) numerical model and a parameter acquisition method. On this basis, an improved interruption overvoltage estimation method is proposed. This method includes an improved MOV numerical model and an improved swarm intelligence algorithm (ISIA), and the proposed ISIA has the advantage of the Levenberg–Marquardt algorithm (LM) and the swarm intelligence algorithm (SIA). The study results, based on the actual MDCCB interruption test whose rated voltage is 10 kV, show that the proposed ISIA can effectively obtain the optimal parameters of the improved MOV numerical model, and this numerical model reflects the relationship between the interruption overvoltage and current of MDCCB well. Compared with traditional MOV numerical models and parameter acquisition methods, the proposed method achieves a maximum absolute error between the estimated and actual interruption overvoltage of only 0.30 kV during the MDCCB interruption period.