Dual-objective optimisation model for the preventive maintenance interval of modular multilevel converter considering random events

Abstract During the operation of the modular multilevel converter (MMC), various random events may affect system reliability and potentially lead to system shutdown. To address this, we propose a dual-objective optimisation model for MMC preventive maintenance intervals that accounts for random events. Firstly, the steady-state probability distribution matrix of MMC operating states is constructed using a semi-Markov process. Simultaneously, a novel cumulative damage operator and improvement factor are introduced to quantify the transient and permanent impacts of random events on reliability. Secondly, a dual-objective optimisation model is established to simultaneously minimise the maintenance cost rate and maximise the energy availability rate, subject to reliability constraints that account for the impact of random events. Thirdly, an improved multi-objective dragonfly algorithm is applied to determine the optimal preventive maintenance interval and overcome local optima. Finally, the proposed model is applied to the Zhangbei flexible direct current Project in Hebei Province, China, to determine the optimal maintenance interval. Compared with the optimisation model that does not consider random events, the mean squared error is reduced by 7.38%. Furthermore, compared to the current fixed-interval engineering strategy, the optimised schedule achieves a cumulative cost reduction of 15.0% over the system lifecycle. The research results provide a theoretical basis for supporting reliable operation and maintenance decisions for MMC.