Research On IGBT Drive Protection And Fault Identification In Inverter

In industrial production,energy conservation and environmental protection have always been the focus of attention.Inverters with excellent speed regulation performance and energy saving effects have been widely used in recent years.However,the working scenario of the inverter is often harsh,which causes the inverter to be prone to failure,and due to the inverter's own technical reasons,problems such as unreliable driving are also prone to occur.Once the inverter fails,it will affect production in a small way and endanger personal safety in a serious way.As the "heart" of the inverter,whether IGBT can be protected in time when a fault occurs and whether it can be reliably driven during operation has become the key to research.This article starts from the three states of the inverter before running,the fault during the running,and the fault,and then studies the drive protection and fault identification of the IGBT.First of all,this paper designs a driver circuit based on the ACPL-T350 chip and makes improvements to effectively suppress the Miller effect during IGBT operation.At the same time,with this drive circuit,IGBT protection circuits,current sampling circuits,and switching power supply circuits are designed to protect the IGBTs in the operating state.Afterwards,in view of the fault situation of IGBT running under short circuit or ground state,this paper designed a set of protection schemes,using special operating strategies to reliably protect IGBT from damage.Then,the Simulink toolbox of MATLAB is used to establish a variable frequency inverter control model,and the simulation model is used to simulate the types of faults that may occur during the operation of the inverter.Finally,the eigenvalues extracted by short-time Fourier analysis are used as the input data of the neural network,and the fault type is used as the output of the neural network to construct a BP neural network model.The neural network is used to locate the fault and solve the complex relationship between the fault characteristics and the fault result.In order to obtain better diagnostic performance,and for the problem that the BP neural network is prone to fall into a local optimum and the recognition accuracy is not high,this paper uses an artificial bee colony algorithm to optimize the fault diagnosis model.Test results show that the optimized model has higher accuracy.