| Wind energy is one of the most promising renewable energy sources.The blade of wind turbine is the core component to realize the conversion of mechanical energy to electrical energy.Its good structural performance is the premise to ensure the long-term stable and efficient operation of the unit.In the long-term operation,the blade will have different degrees of crack phenomenon,which will lead to potential safety hazards.If it cannot be monitored in time,it will bring irreversible damage to the unit.Therefore,the condition monitoring of fan blades has important theoretical significance and engineering application value.Firstly,this paper analyzes the structure of fan blades and common blade detection technologies.Aiming at the shortcomings of existing detection technologies,a fan blade monitoring system based on fiber grating is designed.The work and research are mainly focused on the following aspects:(1)The fan blade model made of glass fiber composite material is constructed.Analyze the mechanical performance of the blade under the rated speed,maximum speed and extreme weather load conditions,and determine the key stress areas of the fan blade through modal analysis and transient dynamic analysis,providing reference for the layout of fiber grating sensors.(2)A fiber grating demodulation system based on scanning laser is built,and the multi-channel fiber grating sensing data is collected and demodulated using "FPGA+STM32" structure.The system first switches the multi-channel photoelectric sensing data through the multi-channel switch CD4051 chip,uses the AD9226 chip for digital-to-analog conversion,and uses the FPGA processing unit to demodulate the peaks of the collected data.The demodulated peaks are transmitted to the STM32F407 chip through the SPI interface,and then the Lo Ra module and the 4G communication module transmit the signals to the upper computer.(3)In practical engineering applications,the reflection spectrum of fiber grating is not strictly symmetrical,which leads to some errors in peak value finding.In this paper,an improved peak-finding algorithm based on asymmetric Gaussian model is proposed.The original spectral data is processed by the five-point moving average filtering method,the peak value of the reflection spectrum is roughly located by the three-point peak-finding algorithm,and the peak position is corrected by the asymmetric Gaussian model.The experimental results show that the algorithm can effectively improve the peak-finding accuracy.(4)The field debugging was carried out for the 1.5MW large test unit,the portable prototype of the monitoring system was built,and the fiber Bgrating was laid.The test results show that the system can meet the stress monitoring requirements of fan blades in complex environments,and has certain engineering value. |
