The Research And Development Of Wind Generating Set Gearbox Fault Diagnosis Instrument Based On BP Neural Network

As the energy and environmental problems have become increasingly prominent, to develop vigorously renewable clean energy has become a consensus. As a kind of renewable clean energy, wind energy is abundant in our country, and also government support for wind power generation is gradually increasing. The wind generating set working in harsh environments is prone to failure, so how to ensure the stable and efficient operation of the wind generating set is an important subject. The gearbox as a double-fed asynchronous generator’s speed-increasing mechanism is the component more prone to failure.Vibration analysis is one of the main technical means currently for rotating mechanical fault diagnosis. The fault diagnosis to wind generating set gearbox is the process of analyzing and processing vibration signal. Embedded system has the advantages of small volume, low power consumption, high performance and high reliability, so it is widely used in all walks of life.According to combination of the theories and methods of fault diagnosis and embedded technology, this thesis develops an intelligent fault diagnosis instrument for wind generating set gearbox. The main contents include the following aspects:First of all, the study and simulation experiment of fault diagnosis theory and related algorithms is done. On the basic principles of the mechanical fault diagnosis, it analyses the mechanism of mechanical vibration of wind generating set gearbox. It theoretically proves the feasibility of this fault diagnosis method by using wavelet packet to decompose vibration signal and extracting energy feature of frequency band, then using the nonlinear mapping ability of BP neural network to establish the nonlinear mapping from fault feature space to the state space. The simulation experiment results further verify the validity of this method.Secondly, based on the requirement analysis of the wind generating set gearbox’s fault diagnosis, it respectively puts forward the general design scheme of hardware and software.Finally, according to the general design scheme, hardware and software are respectively designed in detail. The hardware design is based on ARM microprocessor S3C2440 as the core and then the peripheral expansion circuit is designed. The software design is based on the built basic software platform, including the compile and transplant of the Boot-loader, Linux kernel and root file system. Then it carries on the development of relevant driver program and the design of GUI application program.