Photovoltaic Inverter System Simulation Design And Fault Diagnosis Method

With the worldwide environmental governance issues and the problem of energy reserves has become increasingly prominent, the rational development and efficient use of new energy industries has become the most important issue of the energy sector. Among them, the solar power technology has been the focus of a new clean energy research, after the APEC meeting held will be further developed. PV inverters are the main components of the PV system,change the direct current into three-phase alternating current, has played an important role in connecting the solar array and the load or grid. However,domestic and international research on photovoltaic inverter reliability is still in its infancy, the PV inverter main circuit fault locator design lacks of in-depth discussion, to solve this problem, this paper proposes a fault detection method.This paper focuses on the functions and characteristics of Inverter, from the most basic characteristics of the PV array output, proposed effective photovoltaic inverter system fault diagnosis. The main contents of this paper are as follows:(1)First of all, this paper analyzes the current research of photovoltaic inverter main circuit fault diagnosis, and proposed a diagnostic scheme to be used.(2)Then, in order to facilitate failure analysis, this paper build a DC chopper circuit and select the maximum power point tracking algorithm after study the output characteristics of the photovoltaic array. At the same time, build a model of the inverter control circuit, thereby in MATLAB/Simulink environment construct the PV inverter overall simulation model, and verified its performance fully.( 3) Based on the overall PV inverter simulation model, calculated the inverter output current by wavelet decomposition. Verify the traditional wavelet method advantages and disadvantages, then proposed improved fault feature extraction method, combined with three-phase current high-frequency and low-frequency component to obtain fault characteristic vectors. Lay the foundation of fault identification, ensure automatic positioning the failure point.(4)Set the specific parameter of BP neural network, select the training and testing samples from the obtained fault feature vector group, BP neural network was trained and verified, achieve the PV inverter fault automatically and accurately locate targets. Finally, apply the positioning method in light intensity changes situation, to verify the feasibility.After several simulation experiments confirmed, the proposed photovoltaic inverter fault diagnosis method is correct and feasible, due to retain thelow-frequency information and the high-frequency information solve the disadvantage that cannot be accurately locate which bridge arm failure in a phase, avoid wrong location caused by information missing, has a high diagnostic accuracy and fast convergence, can accurately and efficiently locate solar power inverter main circuit common fault, and can be further expanded to online diagnosis.In addition, the PV inverter fault diagnosis method designed in this paper can also be applied in other areas such as Motor drive, inverter, power supply design. And it has some theoretical research value and practical significance.