基于小波变换的顺风向非平稳风速谱研究

发布时间：2018-04-27 10:22

本文选题：非平稳随机过程 + 功率谱　；参考：《哈尔滨工业大学》2014年硕士论文

【摘要】：现行的结构抗风设计研究中大都是将风速假定为平稳随机过程。通过实测得到的风速数据中成分较为复杂，部分的风速数据为非平稳随机过程，特别是在地形复杂的场地以及台风、飓风等恶劣气候下，风的非平稳性表现的更加明显。由于现行的风工程中没有考虑到非平稳风的影响，所以在抗风设计中会有一定的偏差。本文的主要目的是对风速的非平稳性进行分析，并且提出小波变换估计风速功率谱的方法。本文分析重庆某山区实测风速数据样本，得到脉动风的非平稳特性。通过研究实测山区复杂地形下脉动风的非平稳特性，提出风速非平稳模型的研究方法。并且，利用小波变换得到了非平稳模型下的风速功率谱估计的方法。本文基于进化谱理论，用谐波叠加法模拟生成平稳脉动风速时程，然后选用调制函数进行非平稳脉动风的数值模拟。为了验证本文所模拟的非平稳风速时程的准确度，，本文采用了短时傅里叶变换对风速谱进了估计，并且与理论值对比。为了得到提取趋势项时的最优小波基函数，本文选取不同的调制函数模拟生成非平稳脉动风速，使用sym小波簇，db小波簇，coif小波簇基函数，进行离散正交小波逆变换来提取非平稳脉动风趋势项，并且对比各小波提取的精度，得到了提取趋势项的最优小波基函数。对实测的风速数据进行数据质量控制，剔除坏点，并对该数据进行非平稳性检验，将非平稳样本去掉趋势项后再进行非平稳检验，从而得到了实测山区复杂地形脉动非平稳特性。本文推导出小波变换系数与传统功率谱概念之间的转化关系，并且使用小波变换的方法估计模拟的非平稳脉动风速功率谱。将计算得到的功率谱与理论值进行对比，验证了小波变换估计时变功率谱的准确性。然后，利用小波变换法对实测的非平稳风速数据进行脉动风功率谱估计，本文还提出了基于小波变换的非平稳脉动风功率谱的分析方法。
[Abstract]:The wind speed is mostly assumed to be a stationary random process in the current research on wind resistance design. The components of the measured wind velocity data are more complex and some of the wind speed data are nonstationary random processes, especially in the complex terrain, typhoon, hurricane and other bad weather, and the non-stationary performance of the wind is more obvious. In the current wind engineering, the effect of non-stationary wind is not taken into account, so there will be a certain deviation in wind resistance design. The main purpose of this paper is to analyze the non stability of the wind speed and to propose a method of estimating the wind power spectrum by the wavelet transform.
In this paper, the data samples from the measured wind velocity in a mountain area in Chongqing are analyzed, and the non stationary characteristics of the fluctuating wind are obtained. By studying the non-stationary characteristics of the fluctuating wind in the complex terrain of the mountainous area, the research method of the non stationary wind speed model is proposed. And the method of estimating the wind power spectrum under the non-stationary model is obtained by using the wavelet transform.
Based on the theory of evolutionary spectrum, this paper simulates the generation of stationary pulsating wind velocity with the harmonic superposition method, and then uses the modulation function to simulate the non stationary pulsating wind. In order to verify the accuracy of the simulated non-stationary wind speed time history, the short time Fourier transform is used to estimate the wind velocity spectrum and compare with the theoretical value.
In order to obtain the optimal wavelet basis function of the extraction trend, this paper selects the different modulation functions to simulate the non stationary pulsating wind speed, uses the sym wavelet cluster, the DB wavelet cluster, the coif wavelet cluster function, and carries out the discrete orthogonal wavelet transform to extract the non stationary fluctuating wind trend term, and compares the extraction precision of each wavelet, and obtains the extraction. The optimal small wave basis function of the trend item is used to control the data quality of the measured wind speed data, eliminate the bad points, and check the nonstationary of the data. The nonstationary test is carried out after the non stationary sample is removed, and the non stationary characteristic of the measured terrain fluctuating is obtained.
In this paper, the transformation relation between the wavelet transform coefficient and the traditional power spectrum concept is derived, and the wavelet transform method is used to estimate the power spectrum of the simulated non stationary pulsating wind speed. The calculated power spectrum is compared with the theoretical value, and the accuracy of the time-varying power spectrum estimation is verified by the wavelet transform. Then, the wavelet transform method is used to verify the accuracy of the power spectrum. The fluctuating wind power spectrum is estimated based on the measured non-stationary wind speed data. In addition, a power spectrum analysis method based on wavelet transform for non-stationary fluctuating wind is proposed.

【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU352.2

【相似文献】