Research On Predictive Control Strategy Of Three-Phase Four-Leg Inverter

As a key link of the distributed system,the inverter directly affects the stability and reliability of the system.Inverters operating off-grid usually face severe unbalanced load problems.Compared with countermeasures such as split capacitors and midpoint transformers,three-phase four-leg inverters have the advantages of flexible control,strong anti-unbalanced load capacity,and light structure,which have been paid attention and researched by many scholars.Model predictive control is a relatively novel algorithm.With the rapid development of microprocessors in recent years,its applications in the field of power electronics have gradually increased.It can effectively deal with multiinput and multi-output system,and control the three-phase four-leg inverter as a whole,without the need to deal with the coupling problem of the circuit.So it can avoid the adverse effects such as the decrease in control performance caused by insufficient decoupling in the traditional control,and give full play to the anti-unbalanced load capability of the inverter.This paper starts from the expectation of improving the output power quality of the three-phase four-leg inverter when the load is unbalanced,and first establishes mathematical model of the inverter in the static coordinate system and the rotating coordinate system,and builds prediction model of four-leg inverter on the basis of the state space equation.Secondly,based on the established predictive model,derivation and simulation analysis of the currently used finite set predictive control methods are carried out to verify the correctness of the model and the effectiveness of predictive control.Aiming at the problem that the switching frequency of finite set predictive control is not fixed,which affects the output power quality,a model predictive control method based on continuous set is proposed,which combines carrier-based modulation(CBPWM)to achieve closed-loop control of the system,reducing the amount of calculation and improving output power quality,and the feasibility of this control method is verified based on simulation experiments.Furthermore,from the perspective of control timing,the delay problem of the DSP digital control system and its impact on system stability are analyzed,and the corresponding solutions are discussed from the perspectives of system model and control method.Simulations verify the effectiveness of the measures.In addition,the parameter design of the inverter filter link is carried out,and its influence on the steady-state and transient performance of the system is discussed from the two aspects of inductance and capacitance.Besides,theoretical analysis and simulation verification of the influence of the neutral inductance on the neutral line current are carried out,which provides a theoretical reference for the design of the actual four-leg inverter system.In order to actually verify the feasibility of the proposed continuous set predictive control method,a three-phase four-leg inverter experimental platform is built.Based on the experimental platform,completed the steady-state performance experiment of balanced load,unbalanced load and single-phase load,and the system transient experiment of load switching.At the same time,the influence of the neutral inductance is verified by corresponding experiments.The results show the feasibility of applying the proposed continuous set predictive control method to a three-phase four-leg inverter.