Based on the structure of chute - feed and autoleveHer, an analysis of their working principle and the verification of their practical production results have been carried out. Finally, the future investigation direet...Based on the structure of chute - feed and autoleveHer, an analysis of their working principle and the verification of their practical production results have been carried out. Finally, the future investigation direetiom of chute - feed and card autuleveller are put forward.展开更多
针对滚动轴承故障诊断中样本分布不均衡引起的模型泛化能力差、诊断精度低的问题,从两个方面展开研究:(1)故障样本增广,提出结合变分自编码器(VAE)和生成对抗网络(GAN)的VAE-GAN样本增广模型;(2)改进分类算法,提出基于焦点损失(FL)和卷...针对滚动轴承故障诊断中样本分布不均衡引起的模型泛化能力差、诊断精度低的问题,从两个方面展开研究:(1)故障样本增广,提出结合变分自编码器(VAE)和生成对抗网络(GAN)的VAE-GAN样本增广模型;(2)改进分类算法,提出基于焦点损失(FL)和卷积神经网络(CNN)的FLCNN(focal loss and convolutional neural network)样本分类模型。在此基础上,将VAE-GAN和FLCNN融合,构建VAE-GAN+FLCNN轴承故障诊断模型。首先,将样本量少的故障类输入VAE-GAN模型,通过交替训练编码网络、生成网络和判别网络,学习出真实故障样本的数据分布,从而实现故障样本的增广;然后用增广后的数据样本训练FLCNN分类模型,完成轴承故障识别。试验对比结果表明,所提方法能够有效提升样本不均衡条件下的轴承故障诊断效果,拥有更高的Recall值和F1-score值。展开更多
To investigate the effects of 4-methyl-2-{[4-(toluene-4-sulfonyl)-thiomorpholine-3-carbonyl]-amino}-pentanoic acid isopropyl ester(HD5-6)against cerebral ischemia in rodents,the models with global and focal ischemia w...To investigate the effects of 4-methyl-2-{[4-(toluene-4-sulfonyl)-thiomorpholine-3-carbonyl]-amino}-pentanoic acid isopropyl ester(HD5-6)against cerebral ischemia in rodents,the models with global and focal ischemia were induced by bilateral common carotid artery occlusion plus hypotension(BCAOH)and permanent cerebral artery occlusion(p-MCAO)in mice(n=10–12 per group in BCAOH;n=8 per group in p-MCAO)and rats(n=10-11 in each group).HD5-6 prolonged lifetime and improved neurological function.Neurological deficits score in HD5-6(30 mg/kg)decreased significantly.Malonaldehyde(MDA)in HD5-6-treated mice with ischemia considerably dropped.The infarction volume of the HD5-6-treated rats with MCAO-induced ischemia decreased significantly in the high dose group(P<0.05,i.g.and P<0.01,i.v.).Immunohistochemistry showed that Brain derived neurotrophic factor(BDNF)in the ipsilateral hemisphere increased and Vascular endothelial growth factor(VEGF)decreased with HD5-6 treatment.HD5-6 has protective effects against experimental cerebral ischemia in rodents and the action mechanism may involve anti-oxidation and neurogenesis.展开更多
基于Python语言,结合“老师-学生”模型以及卷积自编码网络提出了Sound Auto Encoder算法,深入研究学生学习算法和实践研究,通过无监督特征学习的方式处理音频数据,根据自编码网络、卷积神经网络原理及相关工作,提出SoundAutoEncoder模...基于Python语言,结合“老师-学生”模型以及卷积自编码网络提出了Sound Auto Encoder算法,深入研究学生学习算法和实践研究,通过无监督特征学习的方式处理音频数据,根据自编码网络、卷积神经网络原理及相关工作,提出SoundAutoEncoder模型及SoundNet中的“老师-学生”模型.在音频数据的特征学习中,通过分析现有算法优势及局限,对SoundAutoEncoder算法模型网络结构及网络学习算法进行分析.通过实验,对算法Sound Auto Encoder和Sound Net进行对比,在五折上,Soimd Auto Encoder算法取得的结果要比Sound Net明显好;在低于600次迭代中,Soimd Auto Encoder算法的结果相较好,Sound Net算法基本处于训练初级阶段,且与可获得的最好结果尚有差距.展开更多
心律失常是常见的心血管疾病之一,目前很多方法通过计算机辅助系统对心电图进行分析以识别心律失常,但由于大多数心律失常数据样本较少,计算机辅助系统识别心律失常效果不佳.本文提出了一种基于混合时频域分析特征提取的卷积神经网络方...心律失常是常见的心血管疾病之一,目前很多方法通过计算机辅助系统对心电图进行分析以识别心律失常,但由于大多数心律失常数据样本较少,计算机辅助系统识别心律失常效果不佳.本文提出了一种基于混合时频域分析特征提取的卷积神经网络方法,该方法提取心电图的RR间期时域特征、希尔伯特-黄变换提取的频域特征和连续小波变换提取的时频域联合特征,经过特征融合后输入卷积神经网络训练分类模型,并采用Focal Loss作为网路的损失函数,实现对心律失常的分类.本文使用MIT-BIH(Massachusetts Institute of Technology-Boston’s Beth Israel Hospital)心律失常数据库验证本文提出方法对4类心电数据分类的结果,实验结果表明,与现有的分类算法相比,本文所提出的混合时频域特征方法能有效提升心律失常分类的准确性.展开更多
文摘Based on the structure of chute - feed and autoleveHer, an analysis of their working principle and the verification of their practical production results have been carried out. Finally, the future investigation direetiom of chute - feed and card autuleveller are put forward.
文摘针对滚动轴承故障诊断中样本分布不均衡引起的模型泛化能力差、诊断精度低的问题,从两个方面展开研究:(1)故障样本增广,提出结合变分自编码器(VAE)和生成对抗网络(GAN)的VAE-GAN样本增广模型;(2)改进分类算法,提出基于焦点损失(FL)和卷积神经网络(CNN)的FLCNN(focal loss and convolutional neural network)样本分类模型。在此基础上,将VAE-GAN和FLCNN融合,构建VAE-GAN+FLCNN轴承故障诊断模型。首先,将样本量少的故障类输入VAE-GAN模型,通过交替训练编码网络、生成网络和判别网络,学习出真实故障样本的数据分布,从而实现故障样本的增广;然后用增广后的数据样本训练FLCNN分类模型,完成轴承故障识别。试验对比结果表明,所提方法能够有效提升样本不均衡条件下的轴承故障诊断效果,拥有更高的Recall值和F1-score值。
基金This work is supported by a grant from the national science and technology major project,a candidate drug for neurodegenerative diseases targeting FKBPs,(2009ZX09103-024).
文摘To investigate the effects of 4-methyl-2-{[4-(toluene-4-sulfonyl)-thiomorpholine-3-carbonyl]-amino}-pentanoic acid isopropyl ester(HD5-6)against cerebral ischemia in rodents,the models with global and focal ischemia were induced by bilateral common carotid artery occlusion plus hypotension(BCAOH)and permanent cerebral artery occlusion(p-MCAO)in mice(n=10–12 per group in BCAOH;n=8 per group in p-MCAO)and rats(n=10-11 in each group).HD5-6 prolonged lifetime and improved neurological function.Neurological deficits score in HD5-6(30 mg/kg)decreased significantly.Malonaldehyde(MDA)in HD5-6-treated mice with ischemia considerably dropped.The infarction volume of the HD5-6-treated rats with MCAO-induced ischemia decreased significantly in the high dose group(P<0.05,i.g.and P<0.01,i.v.).Immunohistochemistry showed that Brain derived neurotrophic factor(BDNF)in the ipsilateral hemisphere increased and Vascular endothelial growth factor(VEGF)decreased with HD5-6 treatment.HD5-6 has protective effects against experimental cerebral ischemia in rodents and the action mechanism may involve anti-oxidation and neurogenesis.
文摘基于Python语言,结合“老师-学生”模型以及卷积自编码网络提出了Sound Auto Encoder算法,深入研究学生学习算法和实践研究,通过无监督特征学习的方式处理音频数据,根据自编码网络、卷积神经网络原理及相关工作,提出SoundAutoEncoder模型及SoundNet中的“老师-学生”模型.在音频数据的特征学习中,通过分析现有算法优势及局限,对SoundAutoEncoder算法模型网络结构及网络学习算法进行分析.通过实验,对算法Sound Auto Encoder和Sound Net进行对比,在五折上,Soimd Auto Encoder算法取得的结果要比Sound Net明显好;在低于600次迭代中,Soimd Auto Encoder算法的结果相较好,Sound Net算法基本处于训练初级阶段,且与可获得的最好结果尚有差距.
文摘心律失常是常见的心血管疾病之一,目前很多方法通过计算机辅助系统对心电图进行分析以识别心律失常,但由于大多数心律失常数据样本较少,计算机辅助系统识别心律失常效果不佳.本文提出了一种基于混合时频域分析特征提取的卷积神经网络方法,该方法提取心电图的RR间期时域特征、希尔伯特-黄变换提取的频域特征和连续小波变换提取的时频域联合特征,经过特征融合后输入卷积神经网络训练分类模型,并采用Focal Loss作为网路的损失函数,实现对心律失常的分类.本文使用MIT-BIH(Massachusetts Institute of Technology-Boston’s Beth Israel Hospital)心律失常数据库验证本文提出方法对4类心电数据分类的结果,实验结果表明,与现有的分类算法相比,本文所提出的混合时频域特征方法能有效提升心律失常分类的准确性.
