| Power electronics is a subject that studies the theory,method and application of electric energy transformation.It is widely used because it can provide any form of electric energy efficiently.Compared with two-level inverter,multi-level inverter is a new type of converter which can achieve high voltage and high power output by changing the topology of the converter itself.Among them,three-level NPC inverter has been widely concerned because of its many advantages,which can achieve stable operation of the system.But because of increasing the number of power devices,the stability of the system has been challenged.Based on the research results of predecessors,this paper proposes a fault diagnosis scheme based on wavelet packet analysis and quantum neural network,which is applied to the fault diagnosis of three-level inverter.Based on the experimental platform,a real-time online diagnosis system is built,and the system status is checked through the PC terminal,and a certain result is achieved.The main work is as follows:First of all,based on the working principle of NPC three-level inverter,this paper studies the inverter’s commutation path,analyzes the inverter’s modulation control algorithm,analyzes the possible faults of the inverter,classifies the faults,summarizes the possible types of faults,and through MATLAB Simulink simulation,the voltage waveform of the three bridge arm is analyzed,and the feasibility of this paper is verified from the waveform diagram,which lays the foundation for further fault diagnosis.Secondly,the fault diagnosis based on wavelet packet and quantum neural network is completed.Through the combination of wavelet packet analysis and KPCA,the fault feature extraction is completed,redundant information is reduced,and the original fault information is retained to the greatest extent.For the fault diagnosis algorithm,this paper uses the latest quantum neural network,which can achieve accurate fault classification,and achieve rapid diagnosis by optimizing gradient descent method.DSP + FPGA is built as the experimental platform of the control system to verify,the method adopted in this paper has achieved good fault diagnosis effect,which is suitable for online fault diagnosis of three-level NPC inverter,and also lays the foundation for the follow-up system visualization.Finally,the overall design of the software is established in the upper computer.Through the use of C#,My SQL database and many other software,a real-time online system belonging to NPC inverter fault diagnosis is designed.The system can realize watching the operation status of the inverter,monitoring the inverter remotely through authorization,realizing remote alarm and sending remote alarm in case of inverter fault Send instructions to change the operation state of the inverter,so as to avoid more serious faults.This software system can also save the system log.By checking the history log,we can fully find the potential problems of the inverter. |
