Research Of Finite Control Set Model Predictive Control For Three-Level Grid-Connected Inverter

Three-level inverter is widely used in high-power grid connected inverter because of its low output harmonic and low voltage requirements of switching devices,but its efficiency is closely related to switching loss.In order to reduce switching loss,the switching frequency of inverter is greatly limited.At the same time,the three-level grid connected inverter also needs to control other targets including the neutral point potential and common mode voltage,while the existing control strategy has some limitations.Finite control set model predictive control is widely used in three-level grid connected inverter because it can control multiple control targets at the same time.However,it still has the problem of difficult to adjust the weight coefficient and modeling error.To solve this problem,this paper proposes a model predictive control strategy with parameter identification and without weight coefficient.The main research work and innovation are as follows:(1)This paper studies the finite control set model predictive control,expounds its basic principle in detail,analyzes its effectiveness in reducing switching frequency and its superiority in multi-objective optimal control,and analyzes the problem that its weight coefficient is difficult to set.(2)In view of the problem that it is difficult to adjust the weight coefficient of the finite control set model predictive control in the multi-objective optimal control,this paper studies the hysteretic model predictive control without the weight coefficient,expounds its basic principle,analyzes its advantages over the finite control set model predictive control,and analyzes the stability of the control system and the modeling error to the control system through mathematical derivation The influence of the stability of the system.(3)Aiming at the problem of modeling error in the hysteretic model predictive control,the adaptive linear neural network algorithm is studied,and the hysteretic model predictive control strategy based on the adaptive linear neural network algorithm is proposed.The identification error and stability of the proposed strategy are analyzed by mathematical derivation,which proves that the proposed strategy can effectively identify the parameters of the prediction model.The selection of initial weight vector and step size of the strategy in three-level grid connected inverter are explained.(4)Finally,a 20 k W three-level grid connected inverter experimental platform is built.Through the experimental platform,the proposed hysteresis model predictive control strategy based on adaptive linear neural network algorithm is verified by experiments,which proves the effectiveness of the control strategy.