Research On On-line Monitoring System Of Wear Particles In Wind Turbine Gearbox Based On Optical Method

The gearbox is an important transmission component of the doubly-fed wind turbine.If the potential failure of the gearbox cannot be found in time,with the continuous operation of the turbine,serious wear will occur inside the gearbox,eventually causing the turbine to stop and even cause serious production accidents.Since a certain amount of wear particles are generated when the gearbox is abrasion,its shape is closely related to the operating condition of the turbine.Therefore,by detecting the wear particles in the oil,the gearbox working condition can be understood earlier,which provides valuable for the maintenance of the situation,avoiding the occurrence of turbine failure.Based on the in-depth analysis of wear particles detection technology,aiming at the shortcomings of the existing monitoring technology,and combining with the on-line monitoring technical requirements of wind turbine gearbox wear particles,an online monitoring system based on optical imaging method and embedded wind turbine gearbox wear particles is studied and designed.In this paper,the mechanical wear and wear particles type characteristics of the turbine gearbox are analyzed firstly,and the correlation between the size of the wear particles and the gearbox operating condition is proved;The factors affecting the clarity of the wear particles image were analyzed,and the method of placing the oil pool vertically was proposed to solve the problems of wear particles deposition jumping and self-locking;According to the system design requirements,an oil-based wear particles detection system based on optical imaging method was designed,and the key devices such as oil inlet,sampling oil pool and microscopic imaging were designed and selected.According to the remoteness of the wind field and the high cost of traditional data processing equipment combined with the performance analysis of the processor,an online monitoring system for wear particles based on embedded processor I.MX6Q and GPRS communication chip SIM800C is proposed;The embedded image processing and control circuit was designed according to the actual needs,and the PC-Linux cross-compilation environment and the streamlined ARM-Linux embedded operating system were built.Real-time acquisition,processing and communication of wear particles images were realized by using the embedded platform.In the real-time image acquisition,the V4L2-based image acquisition program is designed to realize the output format of the image YUYV.In the image preprocessing,the fast graying of the image is realized by extracting the Y component of the image and shifting the image;A variety of smoothing filtering methods were compared,and the salt and pepper noise was filtered by median filtering,and the peak signal-to-noise ratio(PSNR)was used to evaluate the sharpness;Contrast adjustment of the wear particles image using the adaptive histogram averaging method for limited contrast(CLAHE),highlighting some defocused wear particles;The recovery process of motion blur and defocused image is studied in detail,accurate identification of blur angle,length and radius using improved Radon transform,differential autocorrelation and DFT dark loop trajectory,image restoration is better achieved by Wiener filtering.In image post-processing,the Otsu algorithm and morphological algorithm are proposed to realize the segmentation and binary filtering of the wear particles background.According to the Canny operator,the edge detection of wear particles is realized,and 13 characteristic parameters of two types of wear particles size and shape are extracted,which provides necessary preparation for describing the characteristics of wear particles.And the wireless communication program based on GPRS is written to realize the reliable uploading of the wear particles data.Finally,the QT framework was used to compile the monitoring interface with TCP network communication,database management,multi-thread response and wear particles statistics and display sub-functions.The wear particles monitoring experimental platform was built,and the gearbox lubricating oil sample was tested.The test results were consistent with the actual operating conditions,and the testing accuracy met the design requirements.