The Fault Diagnosis System Of Wind Power Gear Box Based On DSP And TQWT Sparse Decomposition

Wind power technology has become one hot topic of the energy research field in current society. The harsh environment in which the wind turbine works, and complex mechanical components, which not only increase wind turbine failure frequency, and seriously affect the overall operation of the wind power plants, but also increase the cost of maintenance personnel. On this basis, it is necessary to design and develop a set of intelligent fault diagnosis system of wind power gearbox to diagnose the fault of wind power generator set.In this paper, the failure mechanisms and types of gear box in wind power generator set are introduced and researched, and the transmission device of gear box is the main research object. For the type of fault signal and characteristic signal, we designed and developed wind turbine gearbox fault diagnosis system based on DSP and TQWT signal sparse decomposition for wind turbine gearboxes signal acquisition, and signal process and analysis in the system.The original signals and diagnostic results are respectively stored in SD cards, and the fault features are displayed on the PC terminal, which facilitates the operator to monitor the wind power generator set and maintenance the fault.The system selects Texas Instruments DSP-TMS320F28335 as the main control chip to build hardware platform, which includes signal acquisition and modification module,peripheral circuits of this control chip and the power module device designed and developed.The software includes two parts. On one hand, using the MATLAB software to edit algorithms program codes, design simulation signals for validating the feasibility of algorithm for TQWT resonance signal sparse decomposition. On the other hand, using the C language to edit the codes of the data acquisition, storage and communication in the CCS software, and convert the algorithm of the TQWT resonance signal sparse decomposition of the MATLAB into embedded codes successfully, which can be identified by the DSP. Then using SVM(support vector machine) to classify and identify fault signals. Eventually the entire system isapplied to fault diagnosis of wind farms for verifying the reliability and effectiveness.