Research On Control Algorithm Of Three-level Inverter Based On Improved Q-MRAS

For high-power and high-performance speed control system,the neutralpoint-clamped(NPC)three-level inverter is widely used in various industries because of its simple structure and superior performance.At the same time,high-performance speed control system often needs speed feedback signal,the use of sensorless control technology will make the installation of the system easier and lower-cost,while its reliability will also be improved.Therefore,this paper focuses on the speed sensorless control based on NPC three-level inverter for the following aspects:Different mathematical models of induction motor are built in different coordinates,the indirect vector control system is designed and its simulation model is built on the platform of Simulink,the simulation results show the designed system has good dynamic and static performance.The basic principle of MRAS speed identification algorithm is studied,and its adaptive law is deduced based on Popov's hyperstability theory.Some problems of MRAS method based on the error of magnetic linkage are pointed out,such as the sensitivity to motor parameters.Based on the analysis of the limitations of the conventional Q-MRAS algorithm,an improved Q-MRAS speed identification algorithm is proposed,by selecting the appropriate reference model and adjustable model,to improve the robustness on the motor inductance parameter.The improved algorithm is studied by small signal analysis,and the stability of the control system is verified by the Routh stability criterion.The speed identification model is established based on the indirect vector control simulation model.The simulation results show that the change of inductance parameters has less impact on the improved algorithm,which also has smaller steady-state error and stronger dynamic tracking ability.The operating principle SVPWM based on NPC three-level inverter is studied,and its neutral-point imbalance problem is analyzed.The distribution factor is used to adjust the working time of small vectors dynamically in order to control the neutral-point.This paper uses three different methods to calculate the distribution factor,whose simulation model are also established.The simulation results show that the neutral-point balance method based on the charge conservation control has shorter adjustment time and better control precision.The experimental platform of sensorless control has been built.TMS320F28335 DSP is used as the master chip to realize the control algo rithm on the CCS5.5.Finally,the experimental results show that the improved Q-MRAS algorithm has better dynamic track performance,meanwhile its accuracy and robustness are also improved.The NPC three-level inverter experimental platform is designed,and the main part of hardware design is analyzed in detail.