滑动移除小波分析法在动力学结构突变检验中的应用
发布时间：2019-07-22 07:03
【摘要】：标度指数计算的即时性与准确性对相关时间序列的动力学结构突变分析至关重要,然而现有方法在即时性与准确性上一直无法兼顾.将小波分析方法与滑动移除窗口技术相融合,提出一种新的动力学结构突变检测方法——滑动移除小波分析法.通过选取不同的滑动移除窗口,分别对构建的线性、非线性理想时间序列进行动力学结构突变分析,结果表明不论是线性时间序列还是非线性时间序列,滑动移除小波分析能够准确地检测到序列的动力学结构突变点及突变区间,对于滑动移除窗口长度依赖性较小,具有很强的稳定性,而且在计算速度上明显优于滑动移除重标极差和滑动移除方差分析方法,将在大数据处理中具有一定的优势.同时分别对线性、非线性理想时间序列添加高斯白噪声,结果表明滑动移除小波分析具有很强的抗噪能力,能够准确地检测到加噪后序列的突变点.对佛坪站日最高温度实测资料的动力学结构突变的准确检测进一步验证了该方法的有效性.滑动移除小波分析法可为具有相关性的系统动力学结构突变的快速、准确检测提供一种途径.
[Abstract]:The timeliness and accuracy of scale index calculation are very important to the sudden change analysis of dynamic structure of related time series. However, the existing methods have been unable to take into account both timeliness and accuracy. A new dynamic structure mutation detection method, sliding removal wavelet analysis, is proposed by combining wavelet analysis method with sliding removal window technique. By selecting different sliding removal windows, the dynamic structure mutation analysis of the constructed linear and nonlinear ideal time series is carried out respectively. the results show that the sliding removal wavelet analysis can accurately detect the dynamic structure mutation point and mutation interval of the sequence, and has strong stability for the length dependence of the sliding removal window. Moreover, the calculation speed is obviously better than that of sliding removal rescaling pole difference and sliding removal variance analysis method, which will have some advantages in big data treatment. At the same time, Gaussian white noise is added to linear and nonlinear ideal time series respectively. the results show that sliding removal wavelet analysis has strong anti-noise ability and can accurately detect the mutation point of noisy sequence. The effectiveness of the method is further verified by the accurate detection of the dynamic structure mutation of the measured data of the daily maximum temperature in Foping station. Sliding removal wavelet analysis can provide a way to detect the sudden change of dynamic structure of the system quickly and accurately.
【作者单位】： 长安大学环境科学与工程学院旱区地下水文与生态效应教育部重点实验室;西安理工大学水利水电学院西北旱区生态水利工程国家重点实验室培育基地;
【基金】：国家自然科学基金青年科学基金(批准号:51409005);国家自然科学基金重大项目(批准号:51190093);国家自然科学基金(批准号:51379014) 中央高校基本科研业务费专项资金(批准号:310829161008)资助的课题~~
【分类号】：N941.3;O211.61
本文编号：2517453
[Abstract]:The timeliness and accuracy of scale index calculation are very important to the sudden change analysis of dynamic structure of related time series. However, the existing methods have been unable to take into account both timeliness and accuracy. A new dynamic structure mutation detection method, sliding removal wavelet analysis, is proposed by combining wavelet analysis method with sliding removal window technique. By selecting different sliding removal windows, the dynamic structure mutation analysis of the constructed linear and nonlinear ideal time series is carried out respectively. the results show that the sliding removal wavelet analysis can accurately detect the dynamic structure mutation point and mutation interval of the sequence, and has strong stability for the length dependence of the sliding removal window. Moreover, the calculation speed is obviously better than that of sliding removal rescaling pole difference and sliding removal variance analysis method, which will have some advantages in big data treatment. At the same time, Gaussian white noise is added to linear and nonlinear ideal time series respectively. the results show that sliding removal wavelet analysis has strong anti-noise ability and can accurately detect the mutation point of noisy sequence. The effectiveness of the method is further verified by the accurate detection of the dynamic structure mutation of the measured data of the daily maximum temperature in Foping station. Sliding removal wavelet analysis can provide a way to detect the sudden change of dynamic structure of the system quickly and accurately.
【作者单位】： 长安大学环境科学与工程学院旱区地下水文与生态效应教育部重点实验室;西安理工大学水利水电学院西北旱区生态水利工程国家重点实验室培育基地;
【基金】：国家自然科学基金青年科学基金(批准号:51409005);国家自然科学基金重大项目(批准号:51190093);国家自然科学基金(批准号:51379014) 中央高校基本科研业务费专项资金(批准号:310829161008)资助的课题~~
【分类号】：N941.3;O211.61
【相似文献】
相关期刊论文 前1条
1 王建平,王竹林,王金玲,罗国军;一种非均匀光照度图像字符二值化小波分析法[J];安徽科技;2005年05期
，本文编号：2517453
本文链接：https://www.wllwen.com/kejilunwen/yysx/2517453.html
