Research On Model Predictive Control Of T-type Three-level Grid-connected Inverter

In order to cope with the energy crisis and follow the low-carbon recycling development model,the development of new energy power generation technology is becoming more and more rapid.As the key equipment for converting new energy into electric energy,grid-connected inverter also develops from traditional two-level inverter to three-level inverter or even multilevel inverter.Among them,T-type threelevel inverter structure inherits the advantages of traditional three-level inverter,and also reduces switching loss,which gradually occupies a place in practical application.However,the traditional PWM strategy will inevitably increase the harmonic content of the grid-connected current of the inverter.Therefore,this thesis uses model predictive control as the grid-connected control strategy of T-type three-level inverter and makes corresponding optimization research.In order to further improve the quality of grid-connected current,this thesis uses LCL filter to reduce the harmonics of gridconnected current and actively damp the corresponding resonance.In this thesis,the structure and operation mode of T-type three-level inverter are introduced in detail,and the problems of neutral point potential imbalance and common-mode voltage of three-level inverter are analyzed,which lays the foundation for further research.Then,this thesis introduces the traditional model predictive control strategy applied to the grid connection of three-level inverter in detail,and puts forward an optimization method of judging sectors and removing some vectors first.By reducing the number of voltage vectors traversed during rolling optimization,the calculation amount of traditional control strategy and common-mode voltage are reduced.In order to balance the neutral point potential on the DC side,a constraint term of neutral point potential and voltage fluctuation is added to the value function.In order to avoid the resonance of LCL grid-connected inverter at this time,an active damping strategy based on capacitor voltage feedback is adopted.Finally,the effectiveness of the improved algorithm is verified by running the simulation model and comparing the control results.Because the traditional model predictive control strategy before and after optimization is a single vector model predictive control strategy,there is a problem of limited control accuracy.In order to further improve the grid-connected control effect and reduce switching loss,this thesis then uses a three-vector model predictive control strategy combined with discontinuous pulse width modulation.Considering the need to ensure the quality and stability of grid-connected power under the actual non-ideal gridconnected conditions,this thesis reconstructs the prediction model through the modeling method of forward and backward differential timing coordination,directly obtains the inverter-side reference voltage needed for modulation from the grid-side reference current,and realizes active damping by extracting the grid-side current harmonics.At the same time,in order to control the neutral point potential balance,the hybrid intermittent pulse width modulation strategy is realized by hysteresis control.The effectiveness and superiority of the proposed control strategy are verified by building a simulation model and running it.Finally,the algorithm proposed in this thesis is verified by experiments on the hardware-in-the-loop simulation platform based on RT-LAB.The experimental results show that the simplified model predictive control strategy achieves better control effect of grid-connected current than the traditional model predictive control strategy.The three-vector model predictive control based on grid-side current can achieve better control effect in both ideal and non-ideal grid-connected situations.This thesis has 96 pictures,8 tables and 83 references.