Modeling And Diagnostic Strategies Study Of Doubly Fed Induction Wind Turbine Failure Based On A Complex Network

For the wind power industry, taking into account the economic benefits, the unitsare in a state of "ill run" in most cases. In the large-scale wind turbine, the variouscomponents, as well as their mutual coupling relations constitute a complex network.Under the role of the network, the components of these "ill run" will affect the normaloperation of other components. In the fault propagation process, those componentsand their mutual coupling relationship through that the fault spread more rapidlyconstitute the weak link in the system. Taking reasonable preventive and protectivemeasures for the weak link in the system can avoid the entire unit outage caused bythe small fault fundamentally. To eliminate the dangerously high risk state, allowingthe system to run in a safe range, the fault propagation characteristics of doubly fedinduction wind turbine should be studied, and the fault propagation paths that canlead to dangerous high-risk can be find out. Thereunder, taking certain preventivemeasures can avoid the system state turning from the initial failure into high-risk,providing the basis for the state of repair, reducing the comprehensive maintenancecosts of the doubly fed induction wind turbine. In addition, according to the faultpropagation model to determine the tightness of the fault contact between the variouscomponents in the system and develop rapid fault diagnosis strategy will help toeliminate system failure, reducing the economic losses caused by the shutdown,reducing downtime, improving economic benefits of wind farms. To study the aboveissues, the main researches in this thesis are as follows:(1) In order to study the propagation characteristics of doubly fed induction windturbine, the fault topology model of doubly fed induction wind turbine is establishedby analyzing the doubly fed induction wind turbine s structure as well as its mainfailure mode. Through statistical analysis of the characteristic parameters of thenetwork, it turned out that the model has small-world properties.(2) The impact of the small-world properties on fault propagation and thecharacteristics of failure transmission in the network is analyzed. Thereunder, the mathematical model of fault propagation path with the highest risk is established. Andthe optimization algorithm is used to solve it, identifying the weak link in the network,obtaining consistent results with the actual operation, which proves the validity of themethod. The results provide an important basis for the state of the doubly fedinduction wind turbine overhaul.(3) Based on the community structure of complex networks theory, a fast faultdiagnosis strategy is proposed. First, the Community structure division of the faulttopology model of the doubly fed induction wind turbine established above is made.Then, the fault trees of the sub-associations are made for fault diagnosis. A faultdiagnosis system was developed to reduce the fault diagnosis time. Provide the basisfor the doubly fed induction wind turbine for rapid troubleshooting and guidance.