Topology Optimization And Fault Diagnosis Methods Of Hybrid MMCs For Offshore DC Wind Turbines

With the rise of power capability and voltage rating for offshore wind farms,the high-power and medium-voltage offshore DC wind turbine,which is suitable for DC connection,is the future development trend.The modular multilevel converter(MMC),with cascaded submodules(SMs),reduces the requirement of voltage rating for power devices.The hybrid MMC(HMMC),consisting of half-bridge SMs(HBSMs)and full-bridge SMs(FBSMs),features fault ride-through capability(FRTC)and becomes a promising topology for the offshore DC wind turbine.However,for topology design,the HMMC cost and reliability are difficult to optimize.For operation control,the accurate and fast SM fault diagnosis is also challenging.This paper focuses on the problems of topology optimization and fault diagnosis,and adopts the 10-MW and10-k V offshore DC wind turbine as an example.With redundancy schemes,the evaluation methods of technical-economic indicators such as power loss,reliability,and cost are investigated.The optimization of original and redundant SM numbers,and fault diagnosis method,which involves SM capacitor voltage ripple suppression strategy,are studied as well.The main research contents of this paper are as follows:(1)In terms of the HMMC power loss evaluation,the switching frequency and voltage stress of power devices change with active redundancy scheme,which result in difficulties of power loss evaluation.Moreover,the power loss model of MMC with single SM structure could not be directly applied to the HMMC.Therefore,analytical power loss models for the HMMC with active redundancy are proposed.Firstly,by considering both fixed-level and voltage-sharing redundancy modes,the phase-shifted carrier PWM strategies of the HBSM and FBSM are analyzed.With different numbers of redundant SMs,switching frequency and voltage stress variations of power devices are revealed.Then,according to the conduction and switching states of power devices,the power loss evaluation models of the HMMC with different redundancy modes are derived.Finally,the impacts of different redundancy modes and SM structures on power loss are verified and compared.(2)In terms of the HMMC reliability evaluation,the temperature and voltage stress variations of power devices with active redundancy scheme,as well as the bypass switch failure rate of passive redundant SM,challenge the reliability evaluation.Thus,analytical reliability models are developed based on the Markov chain.Firstly,according to the power loss calculation method,the SM reliability is modeled with considerations of temperature and voltage stress.Then,since FRTC limits the FBSM proportion,the basic reliability and extra reliability of HMMC are defined,and the HMMC reliability is modeled based on the Markov chain.Finally,by comparing with the conventional reliability models,the feasibility and accuracy of the proposed models are validated.The HMMC reliability with different redundancy schemes is also compared and analyzed.(3)In terms of the HMMC topology optimization,each technical-economic indicator cannot be optimized simultaneously,and the existing methods are not suitable for the complicated and changeable operation conditions.Hence,the multi-objective optimization method for HMMC topology is proposed with wind speed probability distribution,where the objectives are mass,cost,and reliability,and the variables are numbers of original and redundant SMs.Firstly,according to the wind speed distribution,the HMMC mass,initial cost,power loss cost,and reliability are modeled.Then,the objectives with different numbers of original and redundant SMs are analyzed,and the multi-objective optimization method for SM numbers is proposed based on the nondominated sorting genetic algorithm.Finally,with mass,cost,and reliability considerations,the numbers of original and redundant SMs are determined,and the superiority of the proposed method is demonstrated.(4)In terms of the HMMC fault diagnosis,the second-order harmonic circulating current(SHCC)injection method for SM capacitor voltage ripple suppression,and fault diagnosis method immune to SHCC,are presented.The accurate and fast SM fault diagnosis can be achieved.Firstly,the capacitor voltage ripple with SHCC is modeled.With the torque characteristic of wind turbine,the SHCC injection method is proposed to limit the ripple over the entire wind speed range,which facilitates the SM capacitance selection.Then,with phase-shifted carrier PWM strategy,the HMMC circulating current harmonics expressions are derived under HBSM and FBSM fault conditions,based on which the fault diagnosis method is developed.Finally,the effectiveness of the proposed SHCC injection and fault diagnosis methods is verified through simulation and experimental results.The above research achievements can lay theoretical foundation for the evaluation of HMMC technical-economic indicators such as power loss,reliability,and cost,as well as topology optimization,which also provide technical support for the SM capacitor voltage ripple suppression and fault diagnosis of HMMCs for offshore DC wind turbines.