| The environmental pollution is becoming more and more serious today, Countries are paying more and more attention to the utilization of new energy sources, including wind energy. But the wind farm is in poor condition and the units are mostly installed in the tens of meters highly. Causing it can not be found in time and solve the problem effectively when the wind turbine fails. Due to the limitation of Wind Power Field, To replace the internal device until the unit were stopping running. This makes the maintenance cost of the unit becomes higher and the efficiency of power generation becomes lower. And the necessary condition monitoring and fault diagnosis of wind turbine is an effective method to reduce the failure rate and the unit shutdown rate.When the wind turbine is in the working state, Its internal structure in the gear box or bearing will produce a large number of vibration data. It can provide a reliable basis for our fault diagnosis through the analysis of these historical or real time data.Traditional wind turbines are mostly used in stand-alone data processing mode.With the increase of wind turbine vibration data, The speed of data storage and data processing is the bottleneck of the fault diagnosis of wind turbines. In order to solve the problem of single machine mode is difficult to keep up with the requirements of data processing. Since cloud computing was proposed in 2003, its distributed parallel computing and data reliability have been widely concerned. The highly scalable platform Hadoop can store and distribute data to hundreds of inexpensive parallel operation server clusters. To deploy data to the cloud platform for filtering rapidly and analysising efficiently has become the main way of data processing in today's information society.Discrete Fourier transform frequency discretization is implemented by using Fourier transform can realize signal spectrum analysis, the filter frequency response time of computing and signal through the linear system of convolution operation, etc.Therefore, the discrete Fourier transform in the field of signal spectrum analysis plays a very important role. But due to the discrete Fourier transform in time complexity in the process of transformation is too big, can't meet the real-time processing requirements, so later, someone put forward the Courier Fourier transform(FFT). Then,this paper introduces the features of FFT algorithm. Based on the characteristics ofcloud computing technology, We have improved every level operation of the FFT algorithm. The serial FFT algorithm is divided into data completation, indexing operation and butterfly operation in MapReduce, FFT parallel Is realized. Based on the background of the typical fault data of the wind turbine gearbox and bearing of a wind power plant in China, A wind turbine fault diagnosis system is designed in cloud plant.After a full analysis of the requirement, we outline the design principle, topology structure and the main modules of the system, realized the visual interface of the system.At the end of this paper, the performance of the Hadoop platform is tested and analyzed. By comparing the single machine experiment, verify the correctness rate and parallel computing performance of FFT parallelization. Hadoop platform has the characteristics of high real-time performance and high reliability. Demonstrate the superiority of parallel FFT algorithm which based on the MapReduce compared to other solutions. |
PDF Full Text Request