The Control Research Of Three-phase Cascaded Grid-connected Inverter Under Unbalanced Grid Voltage

Under the environment that fossil energy is becoming increasingly scarce,a variety of clean energy comes into human's attention.Grid-connected inverter(GCI)is the interface between clean energy and power grid,which has used widely.The GCI is directly connected to the grid,which will also affect its performance if grid fails.Therefore,it seems to be intensely significant to study the control strategy of GCI under unbalanced grid.The three-phase cascaded grid-connected inverter is easy to be modularized.Compared with Neutral Point Clamped inverter and Flying Capacitor inverter,a large number of diodes and flying capacitor are not needed.Compared with cascaded H-bridge topology,when the number of switch tubes is the same,the output voltage is higher,which can be directly used in high-voltage and high-power applications.Finite control set model predictive control(FCS-MPC)is composed of three parts:prediction model,rolling optimization and feedback correction.It has the advantages of intuitive modeling,easy to understand,control directly and has fast dynamic response.And compared with the traditional inverter control algorithm,it has the advantages of no PWM modulator and parameter adjustment.In this paper,We propose an improved finite control set model predictive control with multi-step prediction(FCS-MPCMSP)algorithm.We use traditional FCS-MPC algorithm in odd cycles and compensate the errors of model by the feedback correction and select the switch states of optimal and suboptimal voltage vector in even cycles.Then we further obtain the optimal switch state from the two states,and make 8 times prediction function calculation into one time voltage calculation,so that the computational complexity of the algorithm can be reduced.Under the unbalanced power grid,the three-phase current is transformed by 3/2,the forward difference,the discrete predictive model and the value function of the grid-connected inverter are established to predict control.According to the four different control requirements,such as,three-phase grid current balance,active power balance,reactive power balance and power-current coordination control.The positive and negative sequencecomponents of the grid voltage are separated by using the positive and negative sequence separation method based on the two-order complex vector filter.Then the grid current reference command is obtained,and the discrete prediction model and the value function are constructed to predictBased on the above theoretical analysis and research,a simulation model is built on the Matlab2014a/ Simulink simulation platform.And the simulation results show that the positive and negative sequence separation method based on the two order complex vector filter can extract the positive and negative sequence components of the grid voltage accurately.When the improved FCS-MPCMSP algorithm is used to control the three-phase cascade inverter,the output current of the inverter can track the reference current very well,reduce the conservativeness of the algorithm,reduce the computational complexity and the operation time of the algorithm.The four different control requirements are achieved under unbalanced grid.