Research On Model Predictive Control Of Hybrid Modular Multilevel Converter

Compared with the traditional modular multilevel converter(MMC),hybrid modular multilevel converter(HMMC)has a very broad application prospect in the field of high voltage direct current(HVDC)transmission,and has received extensive research because of the advantages of self-isolation of DC short-circuit faults.This dissertation takes HMMC as the research object,focusing on model predictive control(MPC)to improve the control performance,and optimizing DC short-circuit fault detection and ride-through control based on the model predictive control.This dissertation expounds the topological structure of HMMC and derives its mathematical model.On this basis,the submodule unified pulse width modulated and capacitor voltages balancing algorithm of sorting are introduced,and their changes during the DC short circuit fault are discussed.In addition,a semi-physical experiment platform based on the RT-LAB real-time simulation system is built,and the platform structure is briefly described.In order to solve the problem that the existing thevenin equivalent model only models the half-bridge and full-bridge submodules,this dissertation establishes a HMMC thevenin equivalent model on the basis of mastering the thevenin equivalent principle.This model integrates half-bridge and full-bridge submodules together for equivalent,that simplifies the model structure and saves simulation time.Finally,the accuracy of the equivalent model is verified by simulation with modulation strategy and voltage equalization method.MPC is the main research content of this dissertation.In order to solve the problems that the control effect of the circulating current control in different applications is not ideal and the current control will affect the AC current,an optimized MPC strategy is proposed on the premise of deducing the HMMC discrete prediction model.The strategy uses the circulating current control parameter?_j to make the circulating current control suitable for MMC systems of various sizes,besides the strategy derives the selection range of?_j theoretically and designs its selection method.In addition,to avoid the influence of circulating current control on AC current,this dissertation uses AC current optimization control to screen the results of the previous stage control.Furthermore,from the perspective of limiting the optimization range and reducing the sorting frequency,the calculation amount of the MPC strategy and the average switching frequency of the switching device are reduced.Simulations and experiments verify the superiority of the control strategy.For solving the problem of DC short-circuit faults,this dissertation discusses HMMC's DC short-circuit fault detection methods and ride-through control based on optimized MPC strategy.This dissertation analyzes the fault characteristics of HMMC bipolar DC short-circuit fault in detail to grasp the change rule of the fault current,on this basis,a method of fault detection using bridge arm current is proposed combined with the optimization of MPC strategy.Compared with the existing detection method,this method does not need to collect the voltage and current information on the DC current-limiting inductor,which reduces the system cost and volume,besides,this method gives a design method of the fault detection threshold.In addition,in view of the shortcomings of the existing fault ride-through control witch using PI controller,this dissertation proposes a DC short-circuit fault ride-through control strategy based on optimized MPC,that simplifies the control method.The simulation results prove that the fault detection method and the ride-through control achieve the design purpose.