对工程振动信号分析与处理中的常用技术进行了研究,包括消除趋势项、数字滤波、频域微积分、生成扫频与白噪声信号、模态参数识别、三分之一倍频程谱、小波分解与重构等,编制了相关程序,并基于实测振动信号进行了分析与处理,与相关商业...对工程振动信号分析与处理中的常用技术进行了研究,包括消除趋势项、数字滤波、频域微积分、生成扫频与白噪声信号、模态参数识别、三分之一倍频程谱、小波分解与重构等,编制了相关程序,并基于实测振动信号进行了分析与处理,与相关商业软件进行了对比校核。消除趋势项可以解决采集数据偏离基线、避免数据无法使用的问题;高通与低通滤波可以根据分析需要任意滤除某频段并获得对应时域数据,从而为振动控制提供理想输入;频域微积分可以通过间接采集加速度或速度信号,分析得到位移信号;扫频和白噪声信号可以为系统振动与隔振性能检验提供输入信号;随机减量法、NExT法、IDT法是系统模态参数识别的重要方法;三分之一倍频程谱是微振动、噪声控制等多领域评价的重要标准;小波分解与重构可以对多源振动信号进行分解,获得各频段成分的各类型振动信号。本研究中总结的若干技术是工程振动控制中最为常用的分析处理技术,对于实际工程应用具有重要的指导意义。The common technologies in the analysis and processing of engineering vibration signals are systematically studied, including the elimination of trend terms, digital filtering, calculus in frequency domain, generation of swept frequencies and white noise signals, modal parameter identification, one-third octave spectrum, wavelet decomposition and reconstruction, etc., and the related programs are compiled, which are analyzed and processed based on the measured vibration signals, which are compared with relevant commercial software. The elimination of trend terms can solve the problem of collecting data deviating from the baseline and avoiding data from being unusable. High-pass and low-pass filtering can arbitrarily filter out a certain frequency band and obtain the corresponding time-domain data according to the analysis needs, so as to provide ideal input for vibration control. Calculus in frequency domain can be used to indirectly analyze the displacement signals by the collected acceleration or velocity signals. The generation of swept frequencies and white noise signals can provide input signals for the vibration isolation performance test. The random reduction method, NExT method and IDT method are important methods for modal parameter identification. One-third octave spectrum is an important criterion for the evaluation of micro-vibration, noise control and other fields. Wavelet decomposition and reconstruction can decompose multi-source vibration signals to obtain various types of vibration signals with a variety of frequency band components. The techniques summarized in this study are the most commonly used analysis and processing techniques in engineering vibration control, which have important guiding significance for practical engineering applications.展开更多
为保证客舱安全,及时识别出旅客的异常行为,基于眼动仪,模拟客舱旅客异常行为,构建试验系统。选取参加过工作实习的空中保卫专业的学生作为被试,获取被试的视觉特征数据。基于多重分形消除趋势波动分析法(Multifractal Detrended Fluctu...为保证客舱安全,及时识别出旅客的异常行为,基于眼动仪,模拟客舱旅客异常行为,构建试验系统。选取参加过工作实习的空中保卫专业的学生作为被试,获取被试的视觉特征数据。基于多重分形消除趋势波动分析法(Multifractal Detrended Fluctuation Analysis of nonstationary time series,MF-DFA),分析航空安全员的视觉搜索特征。结果表明:航空安全员的注视持续时间和扫视幅度与其识别异常行为的能力存在显著的正相关关系;具有高识别能力的航空安全员的注视持续时间和扫视幅度奇异谱宽度均大于低识别能力的航空安全员的奇异谱宽度,具有较强的抗外界干扰能力。将MF-DFA算法引入航空安全员的视觉搜索特征分析,为航空安全员的培训和选拔等提供参考。展开更多
文摘对工程振动信号分析与处理中的常用技术进行了研究,包括消除趋势项、数字滤波、频域微积分、生成扫频与白噪声信号、模态参数识别、三分之一倍频程谱、小波分解与重构等,编制了相关程序,并基于实测振动信号进行了分析与处理,与相关商业软件进行了对比校核。消除趋势项可以解决采集数据偏离基线、避免数据无法使用的问题;高通与低通滤波可以根据分析需要任意滤除某频段并获得对应时域数据,从而为振动控制提供理想输入;频域微积分可以通过间接采集加速度或速度信号,分析得到位移信号;扫频和白噪声信号可以为系统振动与隔振性能检验提供输入信号;随机减量法、NExT法、IDT法是系统模态参数识别的重要方法;三分之一倍频程谱是微振动、噪声控制等多领域评价的重要标准;小波分解与重构可以对多源振动信号进行分解,获得各频段成分的各类型振动信号。本研究中总结的若干技术是工程振动控制中最为常用的分析处理技术,对于实际工程应用具有重要的指导意义。The common technologies in the analysis and processing of engineering vibration signals are systematically studied, including the elimination of trend terms, digital filtering, calculus in frequency domain, generation of swept frequencies and white noise signals, modal parameter identification, one-third octave spectrum, wavelet decomposition and reconstruction, etc., and the related programs are compiled, which are analyzed and processed based on the measured vibration signals, which are compared with relevant commercial software. The elimination of trend terms can solve the problem of collecting data deviating from the baseline and avoiding data from being unusable. High-pass and low-pass filtering can arbitrarily filter out a certain frequency band and obtain the corresponding time-domain data according to the analysis needs, so as to provide ideal input for vibration control. Calculus in frequency domain can be used to indirectly analyze the displacement signals by the collected acceleration or velocity signals. The generation of swept frequencies and white noise signals can provide input signals for the vibration isolation performance test. The random reduction method, NExT method and IDT method are important methods for modal parameter identification. One-third octave spectrum is an important criterion for the evaluation of micro-vibration, noise control and other fields. Wavelet decomposition and reconstruction can decompose multi-source vibration signals to obtain various types of vibration signals with a variety of frequency band components. The techniques summarized in this study are the most commonly used analysis and processing techniques in engineering vibration control, which have important guiding significance for practical engineering applications.
文摘为保证客舱安全,及时识别出旅客的异常行为,基于眼动仪,模拟客舱旅客异常行为,构建试验系统。选取参加过工作实习的空中保卫专业的学生作为被试,获取被试的视觉特征数据。基于多重分形消除趋势波动分析法(Multifractal Detrended Fluctuation Analysis of nonstationary time series,MF-DFA),分析航空安全员的视觉搜索特征。结果表明:航空安全员的注视持续时间和扫视幅度与其识别异常行为的能力存在显著的正相关关系;具有高识别能力的航空安全员的注视持续时间和扫视幅度奇异谱宽度均大于低识别能力的航空安全员的奇异谱宽度,具有较强的抗外界干扰能力。将MF-DFA算法引入航空安全员的视觉搜索特征分析,为航空安全员的培训和选拔等提供参考。
