中文摘要：

针对模块化多电平高压直流输电（MMC-HVDC）系统采用双闭环PI控制参数存在整定难和动态响应慢的问题,提出了一种应用于MMC-HVDC的模型预测控制方法。但模块电容均压排序在具有大规模桥臂模块数的模型预测控制（MPC）策略中存在局限性,提出了希尔排序优化质因子算法（SSPFM）电容均压分组排序方法,通过SSPFM-MPC控制策略,使MMC-HVDC系统具有良好的动态功率响应性能,实现了子模块电容电压均衡与环流抑制之间的良好协调。在PSCAD/EMTDC中搭建了两端401电平MMC-HVDC仿真模型,通过SSPFM-MPC将电压排序的MPC模型的401种组合状态减少到了36种,且系统动态响应良好,验证了所提方法的正确性。

英文摘要：

In view of the problems of difficult parameters setting and slowly dynamic response of double closed-loops PI control adopted by conventional MMC-HVDC system, we propose a model predictive control method applied in MMC-HVDC. Traditional module capacitor voltage-sharing sequencing is limited in large-scale level MPC control strategy. Therefore, SSPFM capacitance voltage-sharing group sequencing is put forward. SSPFM-MPC control strategy endows MMC-HVDC system with favorable dynamic performance, and realizes the perfect combination between sub-module capacitor voltage balancing and circulation inhibition. Finally, two-end 401 level MMC-HVDC simulation model is established in PSCAD/EMTDC. Moreover, 401 kinds of composite states of MPC model of traditional voltage sequencing are decreased to 36 kinds through SSPFM-MPC. The dynamic response of system is favorable, which effectively verifies the correctness of the proposed method.