Research On ADRC Control System Of Modular Multilevel Converter

As the core device to realize power conversion,converter plays an indispensable role in power generation,power transmission and power utilization.With the development of economy,the utilization of power is becoming more and more extensive.The demand for power electronic converters is increasing,and the requirements for power performance indicators are also getting higher and higher.With large-scale renewable energy power generation and the advanced power electronics technologies such as High Voltage Direct Current transmission(HVDC),Flexible AC transmission(FACT),Customized Power,and so on have been widely used in power systems,the development direction of converter becomes higher voltage level and higher power capacity.Converters are based on power semiconductor devices.The voltage level and power level of converters are closely related to the voltage level and power level of power semiconductor devices.Modular Multilevel Converter(MMC)has many advantages such as smooth output waveform,less output harmonic,lower switching loss,simple redundancy and and perfect scalability because of its modular topology.And due to it,MMC has been widely applied into HVDC,power quality control,electric drive and other fields.However,MMC also has some shortcomings,for example,MMC uses the capacitor voltage on the sub-module as the voltage source,when the sub-module is activated,it will charge or discharge the capacitor,so that the capacitive voltage of the sub-module will fluctuate.How to maintain the stability and balance of the capacitive voltage of each sub-module is the primary problem of MMC control system;With the large number of sub-modules,the switching frequency of sub-modules is not fixed when using the Nearest Level Modulation(NLM),and the calculation method of switching loss is complex.How to reduce switching loss and calculate it is a problem that needs to be studied in the heat dissipation design of converters;Due to the wide use of MMC and the complex operation conditions,how to control the circulating current and the capacitor voltage of the sub-module under various disturbances is also a problem to be solved to realize the industrial application of MMC.In recent years,Auto Disturbances Rejection Controller(ADRC)has been widely studied as an effective method to solve the control problems of uncertain nonlinear systems.The core content of this controller is that the ADRC treating the unknown model and disturbances as the total disturbance,to improve the system performance,the Extended State Observer(ESO)is used as a means to real-time estimate the total disturbance of the system and then compensate the total disturbance in the control law.It is proved to be very effective approach to control the uncertain systems with high robustness because it does not sensitive to the parameters and it can estimate and compensate the total disturbance dynamically.Due to it,ADRC is applied in many fields.By introducing the concept of bandwidth,researchers linearization and parameterization the nonlinear form of ESO,Linear Extended State Observer(LESO)greatly reducing the system parameters,make it easier to apply in engineering.Firstly,the average switching model of MMC is established by analyzing its working principle,switching state and energy exchange between input and output.Aiming at the existing control system,a large number of disturbances in the control system are analyzed in detail,and the shortcomings of the existing control system are pointed out.On the basis of existing control systems,the idea of ADRC is further introduced,and a controller with faster response speed and stronger robustness is designed and replace the energy loop controller.In this thesis,the principle of the controller and the tuning method of the parameters are analyzed in detail,and the simulation results under different controllers are given.At the same time,PI + LESO controller is proposed to replace part of PI controller,and the control results are compared and analyzed.Finally,the proposed control system is simulated by using Real Time Digital Simulator(RTDS).The results verify the correctness of the theoretical analysis and the effectiveness of the proposed control system.