Research On Condition Monitoring And Fault Diagnosis System For Double-Fed Asynchronous Wind Turbine

In face of the background of global energy shortage, increasingly serious environment pollution, rising requirements of energy conservation and emissions reduction, as a kind of renewable clean energy, wind power has been paid increasingly attention and acceptance in energy development of world. With a rapid development speed of Chinese wind power industry, the total domestic and new installed capacity are rising year by year. Due to factors of the severe operating environment and the declining of high quality wind resources, the unit designing is developing towards the direction of large capacity, cold resistant in low temperature, high efficiency at low speed. The rapid expansion of wind power industry brings more excess capacity and increasingly remarkable quality problem, and together with the status of "Manufacturing has been paid more attention than management" in wind power industry, the wind power equipment fault occurs frequently and much impact on the unit effectiveness and power grid safety. According to the background and current situation, this paper carries out the research about fault diagnosis of wind turbine, which is based on the knowledge engineering, data mining, signal fusion and mode recognition, etc. The paper can solve the fault mode identification and fault diagnosis, providing technological means for condition monitoring and fault diagnosis system. The research in this paper is as follows.(1) Research on analysis method of wind turbines fault mode and fault knowledge. Starting from wind turbines fault mode attribute analysis, the paper gives symbol identification to those fault modes of different attribute. Using fault tree analysis and failure mode and effects analysis methods, the paper concludes an analysis of hierarchical fault reason, fault impact and specific measures, eventually acquires a comprehensive knowledge of fault information database, which could offer an fault information base for the recognition of fault modes, quantization of fault factors and diagnosis.(2) Research on wind turbines fault modeling and fault feature extraction. Starting from the research of wind turbine fault modeling method, this paper uses the mechanism modeling and inverse modeling to realize the construction of a wind turbine fault model, study in the multivariate dynamic evolution of characteristics of fault development over time scales, and uses the fault feature extraction method under the variable condition to dig out the characteristic information of multiple faults. The work provides a basement of feature information for wind turbines fault quantitative modeling and the intelligent identification of fault modes in different development phases.(3) Research on the method of wind turbines fault mode recognition. Starting from the characteristic analysis of wind turbine fault mode recognition, the paper takes every effort in fault mode intelligent diagnosis, considering different number of characteristics and the coupling characteristics of different measuring points. In the meantime, this research takes the characteristics of fault mode developing with time scale into consideration, solves the intelligent identification problem of fault modes in different severities on the bases of the fault feature of different stages. The research not only forsee the early fault, but also solves the the identification of concurrent fault for wind turbine complex system.(4) Research on fault quantitative analysis and diagnosis method of wind turbines. Based on the result of fault kownledge analysis, the paper realizes the quantitative analysis of wind turbines fault factors by using the method of risk assessment and fault reasons importance calculation combining with wind farm operations experience data. The research result can realize comprehensive diagnosis of fault reasons, fault impact and maintenance measures effectively. At the same time, the results could offer the help to realize the maintenance decision and optimization work.(5) Research on condition monitoring and intelligent fault diagnosis system of wind turbine. Based on the research of core technology theory of condition monitoring and intelligent fault diagnosis methods of wind turbines, this paper develops vibration data acquisition devices for the 1.5MW Sinovel SL 1500 unit, designs the reasonable scheme of data analysis and communication of vibration and SCADA data, and develops the condition monitoring and fault diagnosis system with Delphi 7.0 development platform and Oracal 10.Og database to carry forward technology transfer and engineering applications of this paper.