| As the basis of the bearing member in wind turbine, wind power tower works in long-term bad environment. It is high and heavy. It easily produces defects in long-term serving, which will endanger the safety of tower and the normal operation of wind turbines. These defects need to be detected in time. Also, it is necessary to take remedial measures in time. Otherwise, there will be a major accident of wind power tower.Acoustic emission(AE) technology is a kind of dynamic monitoring technology. It has large monitoring cover area and high-efficiency detection, which can be used in-service detection. To apply AE technology in the monitoring of wind power tower is of importance, which can effectively assess the health of tower. Also, it has a great value in engineering application.Aiming at the complexity in monitoring of wind tower, this thesis extracted the signals collected in monitoring of wind tower by adopting the AE technology. Some results are as follow:(1) The thesis studies the transmission attenuation characteristics of AE signal in wind tower, using amplitude and energy to represent the attenuation property of AE signal. It shows that the weld joint of wind tower has more effect on AE than the distance. The lead-break test shows that the propagation distance of AE signal in the tower is 15 m, but the sensor can receive effective signals far from 6m, which are conductive to the sensor location and explains the application feasibility of AE technology in wind tower. The wavelet analysis of break point and attenuation point signals shows that the energy ratio fell sharply when the signals pass weld line. The energy ratio of low- frequency signal and d3 frequency fell more than other frequency band signals. The frequency of lead-break signals passing welded joint is mainly located in d4 and d3. While, the frequency of not passing welded joint is in d3 and d1, which characterize the signal diversity. Based on statistics of propagation speed of different attenuation conditions, data of field monitoring could be analyzed according to 3700 m/s.(2) Field monitoring experiment of 1.5MW tower barrel with cone type in a wind farm of Dangtang was carried out. Before the experiment, force analysis shows that the middle and lower parts of wind tower bear a larger load, which needs to be supervised. Due to the maximum data size, SPSS Clementine tools can be used to data filtering and extraction, thanks to the most of signals are stable background signals. The amplitude, ringing calculates, duration time and energy data, which are different from those normal data are analyzed. Useless data was filtered out. The abnormal signals are reclassified and restructured. Acoustic emission signals collected through the field test were analyzed by spectrum analysis and wavelet analysis. It showed that the signal frequency of Dj19 wind tower in the 417 seconds of the third test is between 100 kHz to 250 kHz. The major energy concentrated in the d4 and d3 frequencies, and the largest percentage of energy is in d3.The map shows that a part of Dj19 wind tower located at 6.45 m high girth weld in northeast has some suspected defects. |
PDF Full Text Request