精确量化检出大学生的焦虑情绪并对病理因素进行追溯分析,是临床心理治疗和心理危机干预的重要环节,而基于脑电(Electroencephalograph,EEG)信号的深度学习是当前最具发展潜力的一种诊断方法。本研究对传统卷积神经网络(Convolutional N...精确量化检出大学生的焦虑情绪并对病理因素进行追溯分析,是临床心理治疗和心理危机干预的重要环节,而基于脑电(Electroencephalograph,EEG)信号的深度学习是当前最具发展潜力的一种诊断方法。本研究对传统卷积神经网络(Convolutional Neural Networks,CNN)进行改进,提出并构造一个基于“扩展信息输入空间”的神经网络(Neural Network Based on Extended Information Input Space,NN-EIIS)模型,取代CNN末端的分类器;并引入具有独立性的被试对象焦虑量表得分(Score of Anxiety Scale,SAS),作为焦虑情绪量化标准和训练样本集的输出。以某高校大学生为研究对象进行实验,结果表明所提出的方案不仅实现了对焦虑情感的精确量化识别,还能利用所得模型,在一定程度上对大学生焦虑障碍患者的某些重要的内在病理因素进行追溯分析。展开更多
自适应最稀疏时频分析(Aadaptive and Sparsest Time-Frequency Analysis,ASTFA)是一种新的时频分析方法,该方法将信号分解转化为最优化问题,在优化的过程中实现信号的自适应分解。为解决ASTFA方法初始相位函数的选择问题,采用了分辨率...自适应最稀疏时频分析(Aadaptive and Sparsest Time-Frequency Analysis,ASTFA)是一种新的时频分析方法,该方法将信号分解转化为最优化问题,在优化的过程中实现信号的自适应分解。为解决ASTFA方法初始相位函数的选择问题,采用了分辨率搜索改进的ASTFA方法,并进一步结合阶次分析方法提出了基于ASTFA的阶次方法。该方法首先采用改进的ASTFA方法对原始信号进行分解同时获得分量的瞬时幅值,然后对瞬时幅值进行阶次分析从而提取故障特征信息。将该方法应用于变速齿轮传动过程中的时变非平稳振动信号的分析与处理,仿真与实验分析表明该方法能够准确提取变速齿轮的故障特征信息,具有一定的优越性。展开更多
The application of ultrasound techniques to monitor the condition of structures is becoming more prominent because these techniques can detect the early symptoms of defects such as cracks and other defects.The early d...The application of ultrasound techniques to monitor the condition of structures is becoming more prominent because these techniques can detect the early symptoms of defects such as cracks and other defects.The early detection of defects is of vital importance to avoid major failures with catastrophic consequences.An assessment of an ultrasound technique was used to investigate fatigue damage behaviour.Fatigue tests were performed according to the ASTM E466-96 standard with the attachment of an ultrasound sensor to the test specimen.AISI 1045 carbon steel was used due to its wide application in the automotive industry.A fatigue test was performed under constant loading stress at a sampling frequency of 8 Hz.Two sets of data acquisition systems were used to collect the fatigue strain signals and ultrasound signals.All of the signals were edited and analysed using a signal processing approach.Two methods were used to evaluate the signals,the integrated Kurtosis-based algorithm for z-filter technique(I-kaz) and the short-time Fourier transform(STFT).The fatigue damage behaviour was observed from the initial stage until the last stage of the fatigue test.The results of the I-kaz coefficient and the STFT spectrum were used to explain and describe the behaviour of the fatigue damage.I-kaz coefficients were ranged from 60 to 61 for strain signals and ranged from 5 to 76 for ultrasound signals.I-kaz values tend to be high at failure point due to high amplitude of respective signals.STFT spectrogram displays the colour intensity which represents the damage severity of the strain signals.I-kaz technique is found very useful and capable in assessing both stationary and non-stationary signals while STFT technique is suitable only for non-stationary signals by displaying its spectrogram.展开更多
文摘精确量化检出大学生的焦虑情绪并对病理因素进行追溯分析,是临床心理治疗和心理危机干预的重要环节,而基于脑电(Electroencephalograph,EEG)信号的深度学习是当前最具发展潜力的一种诊断方法。本研究对传统卷积神经网络(Convolutional Neural Networks,CNN)进行改进,提出并构造一个基于“扩展信息输入空间”的神经网络(Neural Network Based on Extended Information Input Space,NN-EIIS)模型,取代CNN末端的分类器;并引入具有独立性的被试对象焦虑量表得分(Score of Anxiety Scale,SAS),作为焦虑情绪量化标准和训练样本集的输出。以某高校大学生为研究对象进行实验,结果表明所提出的方案不仅实现了对焦虑情感的精确量化识别,还能利用所得模型,在一定程度上对大学生焦虑障碍患者的某些重要的内在病理因素进行追溯分析。
文摘自适应最稀疏时频分析(Aadaptive and Sparsest Time-Frequency Analysis,ASTFA)是一种新的时频分析方法,该方法将信号分解转化为最优化问题,在优化的过程中实现信号的自适应分解。为解决ASTFA方法初始相位函数的选择问题,采用了分辨率搜索改进的ASTFA方法,并进一步结合阶次分析方法提出了基于ASTFA的阶次方法。该方法首先采用改进的ASTFA方法对原始信号进行分解同时获得分量的瞬时幅值,然后对瞬时幅值进行阶次分析从而提取故障特征信息。将该方法应用于变速齿轮传动过程中的时变非平稳振动信号的分析与处理,仿真与实验分析表明该方法能够准确提取变速齿轮的故障特征信息,具有一定的优越性。
基金Projects(UKM-KK-03-FRGS0118-2010,UKM-OUP-NBT-28-135/2011)supported by FRGS Universiti Kebangsaan Malaysia,Malaysia
文摘The application of ultrasound techniques to monitor the condition of structures is becoming more prominent because these techniques can detect the early symptoms of defects such as cracks and other defects.The early detection of defects is of vital importance to avoid major failures with catastrophic consequences.An assessment of an ultrasound technique was used to investigate fatigue damage behaviour.Fatigue tests were performed according to the ASTM E466-96 standard with the attachment of an ultrasound sensor to the test specimen.AISI 1045 carbon steel was used due to its wide application in the automotive industry.A fatigue test was performed under constant loading stress at a sampling frequency of 8 Hz.Two sets of data acquisition systems were used to collect the fatigue strain signals and ultrasound signals.All of the signals were edited and analysed using a signal processing approach.Two methods were used to evaluate the signals,the integrated Kurtosis-based algorithm for z-filter technique(I-kaz) and the short-time Fourier transform(STFT).The fatigue damage behaviour was observed from the initial stage until the last stage of the fatigue test.The results of the I-kaz coefficient and the STFT spectrum were used to explain and describe the behaviour of the fatigue damage.I-kaz coefficients were ranged from 60 to 61 for strain signals and ranged from 5 to 76 for ultrasound signals.I-kaz values tend to be high at failure point due to high amplitude of respective signals.STFT spectrogram displays the colour intensity which represents the damage severity of the strain signals.I-kaz technique is found very useful and capable in assessing both stationary and non-stationary signals while STFT technique is suitable only for non-stationary signals by displaying its spectrogram.
