Three-Level Active Power Filter Based On Model Predictive Control

With the progress of power electronic technology and the development of manufacturing industry,more and more nonlinear equipment are put into production and manufacturing.The input of these nonlinear loads will bring huge harmonic interference and reactive power impact to the power grid.If not treated,it will seriously affect the power quality of the power grid,and even cause safety accidents and major economic losses.As a new power management device,active power filter（APF）has attracted extensive attention because of its flexible compensation,good dynamic performance and small volume.This thesis takes the diode clamped（NPC）three-level APF topology as the research object,and studies the key technologies in application,such as current model predictive control（MPC）,harmonic current detection,neutral point potential balance,etc.the main contents are as follows:Firstly,this thesis expounds the necessity of harmonic suppression,summarizes the research status of APF topology,harmonic detection,current tracking and DC side capacitor voltage stabilization,analyzes the topology and working principle of diode clamped three-level APF in detail,deduces the mathematical model of the system,and studies the neutral point potential balance mechanism,it is pointed out that the neutral point potential balance at DC side can be realized by controlling small vector.Secondly,the three-level APF control technology is studied.The mathematical derivation of instantaneous reactive power theory is given,and the i_p-i_q method is selected as the harmonic reference current extraction algorithm in this thesis.The model predictive control is studied,the APF system prediction model is established according to its basic principle,the corresponding value function is constructed,and the implementation process of three-level APF based on MPC is given.On this basis,aiming at the problem that there are still many rolling optimization times of the improved MPC,difficult to set the weight of the objective function and large amount of calculation in the traditional MPC algorithm,this thesis proposes an improved MPC algorithm with single objective function,which only takes the reference current and feedback current as the objective function,omits the weight coefficient setting process,and reduces the rolling optimization times from 27 times to 21 times.Aiming at the problem that there are still many rolling optimization times,an improved MPC algorithm of single objective function based on sector judgment is proposed.On the basis of maintaining the advantages of single objective function,the rolling times are reduced to 7 times,which further reduces the amount of calculation and control complexity.Finally,a three-level APF system based on MPC is built on the MATLAB/Simulink software platform.Under different working conditions,the traditional model prediction and two improved MPC algorithms are used for simulation.The results show that the single objective function improved MPC algorithm based on sector judgment proposed in this thesis has better results in harmonic current tracking and capacitor voltage stabilization,the correctness of the proposed control algorithm is verified.In this thesis,the NPC three-level APF experimental platform with the rated system capacity of 10k VA is built for experimental verification.The results show that the single objective function improved MPC three-level APF based on sector judgment is correct and feasible.