| Modular multilevel converter(MMC)adopts distributed structure and has the advantages of modular design,good output waveform quality,low switching frequency of devices,and easy realization of modular redundancy design,which is widely used in high voltage High power converter field.However,MMC usually contains a large number of submodules(SM),and each submodule needs to be equipped with a capacitive voltage sensor to collect the capacitive voltage.So many sensors not only make the cost of the system high,but also make the data acquisition and communication system overburdened.How to achieve stable operation of MMC while reducing or eliminating sensor usage is an important research direction.This paper takes the modular multilevel converter as the research object,studies the optimal operation and control of the modular multilevel converter under the condition of few sensors,and which is funded by the key project of the National Natural Science Foundation of China “Research on the key basic problems of the integrated system of multiphase direct drive permanent magnet wind power generation and conversion(51737004)”.This paper mainly studies the capacitor voltage measurement error reduction of MMC with reduced capacitor voltage sensors,overvoltage protection of submodule of MMC with reduced voltage se nsors and deadbeat control of MMC with reduced sensors under the inductance parameter mismatch of bridge Arm.The specific content as follows:Modular multilevel converters usually have a large number of submodules,and the nearest level modulation(NLM)technology is more suitable for the occasion where the MMC contains a large number of submodules than the carrier modulation technology.Although the existing NLM-based submodule capacitor voltage measurement technology of MMC with reduced sensors can reali ze the operation of the modular multilevel converter with reduced capacitor voltage sensors,it has the problems of large capacitor voltage measurement error and poor capacitor voltage balance.From the perspective of submodule capacitor voltage sampling m odulation,this paper focuses on the analysis of the source of submodule capacitor voltage measurement error in the existing NLM-based submodule capacitor voltage measurement technology of MMC with reduced sensors,and proposes a measurement algorithm to reduce the submodule capacitor voltage measurement error.The proposed new algorithm forcibly keeps the operating state of the submodule in the previous control cycle unchanged when the number of submodules that need to be put in or removed changes to “1” in two consecutive control cycles,and then according to the current direction of the bridge arm of MMC with reduced sensors,select the submodule to switch on or off,and directly calculate the actual value of the capacitor voltage through the arm voltage sensor sampling voltages.The proposed algorithm changes the submodule switching quantity change to “1” event into the state change to“1” event,so as to realize the direct measurement of the sub-module capacitance voltage from a probabilistic event to an inevitable event,thereby improving the submodule capacitance voltage direct measurement probability.With the improvement of the submodule capacitor voltage direct measurement probability,the number of corrections of the observed submodule capacitor vol tage is increased,which finally reduces the measurement error of the submodule capacitor voltage.At the same time,the conditions for the implementation of the proposed algorithm,the probability increase of the actual value of the submodule capacitor voltage,and the influence of the proposed algorithm on the switching frequency and capacitor voltage balance of the operation of the MMC with reduced voltage sensors are analyzed.Finally,simulation and experimental results verify the feasibility and effec tiveness of the proposed algorithm.In order to further reduce the measurement error of the submodule voltage in MMC with reduced sensors system,firstly,taking a phase bridge arm submodules divided into two groups as an example,then,from the perspectiv e of the capacitance voltage measurement topology of the MMC with reduced sensors,a submodule capacitance voltage measurement topology in which the capacitance voltage sensor is cross-stacked to measure the capacitance voltage of a submodule is proposed.Compared with the capacitance voltage measurement topology of the MMC with reduced sensors used in Chapter 2,the proposed topology can not only further improve the probability of the MMC submodule capacitance voltage being directly measured,thereby reducing the measurement error of the submodule capacitance voltage,but also increase the flexibility of bridge arm submodule configuration.Finally,the simulation results show that the proposed topology has a certain superiority in both the traditional submo dule capacitance voltage measurement algorithm and the submodule capacitance voltage measurement algorithm proposed in the second chapter.The capacitor voltage of the submodules of the modular multilevel converter operating under few voltage sensors canno t be directly measured in real time.Most of the submodules capacitor voltages are obtained by the voltage observer,and the overcharging of some submodules whose capacitor voltage is obtained by the capacitor voltage observer will lead to over-voltage on the submodule.In order to solve this problem,a capacitor overvoltage elimination strategy for reduced-voltage-sensor-based MMC is proposed in this paper.The overvoltage elimination is consists of pre-measurement and in-process measurement of capacitor voltage,which can obtain the actual value of submodule capacitor voltage and determine the submodules with overvoltage and overvoltage trends in real time,so that to sort these submodule capacitor voltages in the correct sequence and eliminate the overvoltage that occurred on submodules.Due to the reduction of the number of capacitive voltage sensors,most submodules obtain the submodule capacitive voltage through the voltage observer,and the capacitive voltage observer is greatly affected by the paramet ers.When the parameters of the modular multilevel converter with reduced sensors change,the operation of MMC with reduced sensors could be greatly affected.Existing literatures have analyzed the influence of capacitance parameter offset on the operation of MMC with reduced sensors system,and given an online correction strategy for offset capacitance parameters.However,the influence of bridge arm inductance mismatch on the operation of MMC with reduced sensors and its related policies are not covered.Aiming at the instability of submodule capacitor voltages and output current when the MMC with reduced sensors system uses the traditional control method under the condition of bridge arm inductance mismatch,a dead-beat control strategy for the MMC with reduced sensors is proposed.The control strategy combines the discretization model of the operation of the MMC with reduced sensors,directly takes the output current and circulating current as the control objects to control the MMC with reduced sensors,which is to improve the stability and the operating efficiency of the MMC with reduced sensors system under the condition of bridge arm inductance parameter mismatch.Simulations and experiments demonstrate the effectiveness and feasibility of the proposed method. |