Research On Acoustic Emission Signals For The Damage Of The Fan Blade Based On De-noising Technology And Blind Source

Along with the increasing consumption of conventional energy, new energy-wind energy developrapidly.The regular and efficient operation of WTGS(wind turbine generator system) is the foundationof widespread application of the wind energy. The fan blade is the crucial structure of WTGS. In orderto avoid the accident that while the wind turbine blades are working, the study of the blades damagemechanism and damage pattern is significant meaningful.According to relevant document, the author realized that the two key problems which are usuallyexisted in the structural health monitoring of the fan blade. They are:1. The structure damage signalsgenerated by different source of damage in the process of communication will appear the phenomenonof mixing and overlay each other;2. Acoustic emission signal can contain a lot of background noise.Based on the two key issue, this article focus on the analysis and processing of the acoustic emissionsignal collected in the experiment for the wind turbine test specimens. The main content and work areas follows:1. As the glass fiber composite material is most used in the current production of the windturbine blade,so design and produce a certain size of wind turbine blades composite specimens, andthe specimens are produced in strict accordance with the actual number of layer and fiber. Use theexperimental platform to apply load on the specimen until it damaged and tear, and collected theacoustic emission signal in the whole process.2. According to the acoustic emission signals collected in the experimental conditions, study thecorresponding noise reduction technology. Through the simulation analysis of the effect of wavelettransform modulus local maxima de-noising and wavelet threshold de-noising, use this two kinds ofnoise control schemes to reduce noise of the acoustic emission signal for actual wind turbine blades.contrast the noise reduction effect of two methods, the author select the way which is based on thepenalty strategy of wavelet threshold de-noising pre-treatment techniques on the signal noisereduction.3. Considering the acoustic emission signals produced by the whole tearing failure process of thecomposite laminates are multi-type and multi-mode, and use of the independent component analysisto separate the gathered acoustic emission signals.4. And go ahead with the separated signals to detailed study by computer program which is programmed with MATLAB. Then analysis and identify the damage status by connect with ModalAE.The result affirms that blind source separation technology can be to separate the audibilityacoustic emission signals, and summarized audibility acoustic emission signals in the differentfrequence have their corresponding characters. Through the analysis to the acoustic emission signalsin the audibility zone, it affirms that the wind turbine blades specimens consist of glass fibercomposite have four damage source, they are matrix crack, interface de-bonding, separated layer andfiber breakage.