Research On Model Predictive Conteol Strategy Of DC Microgrid Energy Storage Converter

In order to make renewable energy further lead the mainstream direction of energy revolution and consumption revolution,and realize the great vision of "carbon peak,carbon neutrality",DC microgrid,as a new generation of low-carbon and efficient energy system,has attracted wide attention from the society.Each unit in the DC microgrid is connected to the DC bus through a power electronic conversion device.The energy storage unit is the main force to suppress the random and fluctuating defects of renewable energy generation to achieve " peak load shifting".The selection and control of the energy storage converter connecting the energy storage unit and the DC bus is very important for the DC microgrid.In this paper,the bidirectional Buck/Boost converter and interleaved bidirectional DC-DC converter used in different occasions of energy storage unit in DC microgrid are taken as the research objects,and their control strategies are studied in depth.The traditional double closed-loop control technology of bidirectional Buck/Boost converter is mature,but the design of PI parameters is complex.Finite Control Set Model Predictive Control(FCS-MPC)has the advantages of flexibility,no modulation,multi-objective nonlinear control,etc.,and has become a research hotspot in the field of power electronic converters.The control signal.is realized by a single controller,and the light-on state can be selected from the different working states of all the switches to act on the converter with a preset,minimum cost function and the most expected performance of the system.However,as an openloop structure,the FCS-MPC prediction model module will affect the prediction accuracy due to objective factors.In the case of large disturbances,the cost function in the rolling optimization link will also require the system to run for a longer time in the "non-optimal state" because the adjacent control cycles are not connected.Based on this,this paper proposes a feedback correction model predictive control strategy based on the idea of "two beat prediction".On this,this paper proposes a feedback correction model predictive control strategy based on the idea of"two beat prediction".On the basis of retaining the simple and rapid control process,the FCSMPC control effect is optimized through new prediction model and rolling optimization.For the interleaved bidirectional DC-DC converter used in high-power applications,the FCS-MPC control strategy is used to balance the two-phase inductor current to reduce the input and output current ripple of the energy storage unit.In view of the original defects of FCS-MPC,the interleaved bidirectional DC-DC converter FCS-MPC is optimized,and a new prediction model and a cost function of the rolling optimization link are designed.Based on the characteristics of the interleaved topology,the system performance is optimized,the control complexity is reduced,and the DC bus voltage stability and the two-phase inductor current balance are better realized.At the same time,the simulation and experimental models of the above topology and related control methods are constructed respectively to verify the control effect of FCS-MPC and its optimized control.Finally,it is shown that FCS-MPC has full advantages in the field of converter control.The optimized FCS-MPC can better maintain the DC bus voltage and control the converter to respond faster to energy storage charging and discharging.