Design And Implementation Of Inverter Fault Diagnosis System Based On FPGA

With the development of new energy sources and the rapid expansion of industries such as wind and photovoltaics,inverters,as a key component in the electrical energy conversion process,have received more attention.When the inverter is working,any component failure may cause the quality of the inverter’s output power to decrease,and in severe cases,it may cause damage to the equipment and cause serious economic losses.Therefore,it is necessary to implement inverter fault diagnosis accurately and quickly.Neutral Point Clamped(NPC)three-level inverter is currently the most common circuit topology,which has the advantages of less output harmonics and simple topology.Therefore,this article takes the NPC three-level inverter as the research object,and conducts the following research:(1)Introduced the working principle and fault types of NPC three-level inverters,and analyzed the circuit conduction under different fault conditions.Taking complex faults as the research object,the fault types are divided into 48 types,and determined the type and location of signal acquisition.(2)Design voltage signal acquisition system based on Field-Programmable Gate Array(FPGA).Choose the appropriate AD acquisition module,FPGA control module and data transmission module.Then,used Verilog language to complete the software part writing,and complete the construction of the voltage signal acquisition system.Use the signal generator to output the inverter analog fault signal.The signal acquisition system collects voltage signals under normal and 48 complex fault conditions,and provides data support for subsequent fault feature extraction and fault classification.(3)The load voltage signal is collected,and the complementary ensemble empirical mode decomposition(CEEMD)is performed to obtain a series of intrinsic mode functions(IMF).The sample entropy value of each component is calculated,and the positive halfcycle proportional coefficient of the voltage signal is calculated together to form the fault feature vector.Particle swarm optimization(PSO)support vector machine(SVM)is used for fault classification.The classification accuracy rate reaches 100%,and compared with four commonly used algorithms,the method proposed in this paper has certain advantages.(4)A LabVIEW-based NPC three-level inverter complex fault diagnosis system was developed.Based on the previous algorithm theory,realize the mixed programming of LabVIEW and MATLAB.Combining the simple interface design function of LabVIEW with the excellent numerical calculation ability of MATLAB,realizes the functions of acquisition signal display,decomposition,fault classification,etc.Input different fault signals to test the system,which verifies that the fault diagnosis system developed in this paper has good working performance.