An optimal control set model prediction control method for a modular multi-level converter

Abstract

The invention relates to an optimal control set model prediction control method for a modular multi-level converter. The method comprises the steps of: according to insertion number optimization and avoltage fluctuation range of sub-modules of a modular multi-level converter in a voltage ideal state, determining an optimal control set of the range of the insertion number of the sub-modules, wherein the sub-modules include upper bridge arm sub-modules and lower bridge arm sub-modules; in the range of the optimal control set, performing optimization through a model prediction control method todetermine the optimum insertion number of the sub-modules; according to the optimum insertion number, achieving bridge arm voltage balance of the modular multi-level converter and generating PWM switch signals to achieve model prediction control of the modular multi-level converter. Compared with the prior art, the method has the advantages of fewer optimization times, small calculation quantity and static and excellent dynamic control performance.

Description

technical field [0001] The invention relates to the control field of modularized multilevel converters, in particular to an optimal control set model predictive control method for modularized multilevel converters. Background technique [0002] The modular multilevel converter is a converter with the characteristics of modular structure, easy cascading, and low output voltage and current harmonics. Due to its advantages in loss, redundancy, and switching frequency, it is widely used in high-voltage and high-power engineering sites such as high-voltage direct current transmission, flexible power transmission, wind farm grid connection, and medium-high voltage electric drive. [0003] Because the bridge arms of the modular multilevel converter are connected in parallel at both ends of the DC bus, the capacitors of the sub-modules are suspended, and the fluctuation of the capacitor voltage of the sub-modules makes it difficult for the sum of the voltages of the bridge arms conn...

AI Technical Summary

Problems solved by technology

Existing literature mostly adopts the design of PI, PR or vector decoupling control methods to suppress the circulating current of modular multilevel converters. However, due to the large number of switch combinations and strong nonlinearity, the stability margin of traditional controllers is reduced, and the dynamic Poor performance, which limits the promotion of engineering applications

[0004] Model predictive control has been successfully applied in modular multilevel converters and has attracted widespread attention because of its intuitiveness, flexibility, strong robustness, and adaptability to nonlinear and multi-control objects. The existing model predictive control is often Firstly, the global optimization calculation is performed on the number of submodules that may be inserted into the upper and lower bridge arms, and then the bridge arm voltage balance of the modular multilevel converter is realized according to the determined number of inserted submodules in the upper and lower bridge arms, and PWM is generated. switch signal, so as to realize the model predictive control of the modular multilevel converter. Since this method needs to optimize the number of inserted sub-modules at the global level, the calculation amount is large and the calculation is complicated, which leads to poor engineering practicability. powerful

Images