| With the aggravation of environmental pollution,energy structure imbalance and other contradictions,renewable energy power generation technology represented by solar energy has gradually attracted the attention of scholars at home and abroad.As the core apparatus connecting renewable energy and power grid,the inverter significantly enhances the power transmission quality of the grid-connected system by the reasonable choice of its structure and control strategy.The energy storage quasi-Z-source inverter(ES-qZSI)not only can realize the functions of boost and contravariant in a single stage,but also can flexibly adjust the power vacancy of the system and suppress the impact of photovoltaic power fluctuation on the power grid.This dissertation takes ES-qZSI as the research target,and employs finite control set model predictive c Control(FCS-MPC)strategy to achieves high-performance grid-connected operation of photovoltaic grid connected inverter.This dissertation mainly includes the following aspects:(1)According to the structure of ES-qZSI,its basic working principle is analyzed,and corresponding mathematical models are established in different coordinate systems.Considering the large amount of computation in traditional FCS-MPC algorithm,an improved FCS-MPC algorithm for ES-qZSI is researched.This method realizes the judgment of whether the switching state of the inverter is shoot through(ST)state by constructing the inductor current sub-cost function.If the minimum sub-cost function is achieved in the ST state,the ST state is the optimal state without checking the other seven states,otherwise the optimal switching state needs to be optimized online from the seven states.The simulation results demonstrate that the improved FCS-MPC strategy lessens the computational complexity by about 50% compared with the traditional strategy.(2)In order to improve the current control accuracy of traditional FCS-MPC algorithm,a three-vector model predictive current control(TV-MPCC)based on ES-qZSI is proposed.The algorithm first obtains the duration of each voltage vector by taking the minimum error current value of the direct and quadrature,and then constructs six virtual voltage vectors that can be adjusted in direction and amplitude according to the principle of the nearest three vectors,which effectively increases the current control accuracy of the system.Only by online predicting the current value for six times,optimal voltage vector acting on inverter is obtained.The simulation results show that compared with traditional MPC algorithm,the current THD of this proposed method is reduced by 2.92%,and the grid-connected operation capability of inverter is improved.(3)In order to overcome the disadvantage of high switching losses caused by the unfixed switching frequency of traditional direct power control strategy,a model predictive direct power control(MP-DPC)for ES-qZSI system is proposed.In this control,the instantaneous values of active and reactive power output by the inverter are taken as the controlled variables.By eliminating the tracking error between the instantaneous power and the given power value,the average voltage vector in αβ coordinate is obtained,and the fixed switching signal is generated by combining space vector modulation for Z-Source inverter.It is verified that the simulation results show that the ES-qZSI achieves six shoot-through actions in each switching period to achieve the constant-frequency control of the system,which verifies the correctness of the proposed control strategy. |
