正在加载图片...
·822· 工程科学学报,第41卷,第6期 900 [4]Zhou X F,Yang F,Guo L,et al.Analysis and preventive recom- 训练集 800 验证集 mendation of national confined space accidents due to asphyxiation 700 口测试集 and poisoning from 2014 to 2015.J Eniviron Occup Med,2018, 600 0误差 35(8):735 500 (周兴藩,杨凤,郭玲,等.2014一2015年全国有限空间作业 400 中毒与室息事故分析及预防建议.环境与职业医学,2018,35 300 (8):735) 200 I00 5]Mejias C,Jimenez D.Munoz A,et al.Clinical response of 20 people in a mining refuge:study and analysis of functional parame- ters.Safety Sci,2014,63:204 [6]Selman J,Spickett J,Jansz J,et al.An investigation into the rate 误差值 and mechanism of incident of work-related confined space fatali- 图8误差直方图 ties.Safety Sci,2018,109:333 Fig.8 Error histogram ]Li GJ.Research on Human Heat Tolerance under Extreme Hot, Humid and Lou-xygen Entironment [Dissertaion].Tianjin 进行分类,相较于其他的机器学习方法,深度学习无 Tianjin University,2008 需人工进行特征提取且学习特征更加全面,保证了 (李国建.高温高湿低氧环境下人体热耐受性研究[学位论 识别的准确性. 文].天津:天津大学,2008) (2)通过多次试验得出深层神经网络在具有3 Kamshilin AA,Nippolainen E,Sidorov I S,et al.A new look at 个隐藏层,且每层有9个神经元的情况下分类效果 the essence of the imaging photoplethysmography.Sci Rep,2015, 5:10494 较好,且能保证较快的训练速度. 9) Shin H,Min S D.Feasibility study for the non-invasive blood (3)通过混淆矩阵分析和ROC曲线分析可知 pressure estimation based on ppg morphology:normotensive sub- 深层神经网络对不同氧气浓度的PPG信号分类性 jeet study.Biomed Eng Online,2017,16(1):10 能优良,在训练集中分类准确率可达97.9%,在验 [10]Zhou Q,Tu H,Zuo J X.Blood oxygen saturation real-time moni- 证集中分类准确率可达94.8%,在测试集中的分类 toring system research based on PPG.Inform Res,2017,43(3): 准确率为92.8%.在全部的数据中准确率为 75 96.3%,认为能够准确的识别人体是否处于氧气耐 (张强,涂浩,左佳鑫.基于PPG的血氧饱和度实时监测系 受极限状态 统研究.信息化研究,2017,43(3):75) (4)所搭建的深层神经网络进行低氧状态评估 [11]Njoum H,Kyriacou P A.Photoplethysmography for the assess- ment of haemorheology.Sci Rep,2017,7(1):1406 只需采集2000个离散点(4s采集完成),因而能够 [12]Ma J L,Wang C,Li Z J,et al.Study of measuring heart rate 保证整个评估过程在4s内完成,从而使得作业人员 and respiration rate based on PPG.Opt Tech,2011,37(3):309 暴露在低氧环境中的时间减少,保证了作业人员的 (马俊领,王成,李章俊,等.基于PPG的心率和呼吸频率 安全 的测量研究.光学技术,2011,37(3):309) 03] Tamura T,Maeda Y,Sekine M,et al.Wearable photoplethys- 参考文献 mographic sensors-past and present.Electronics,2014,3(2): [Burlet-Vienney D,Chinniah Y,Bahloul A,et al.Occupational 282 safety during interventions in confined spaces.Safety Sci,2015, [14]Khadse C B,Chaudhari M A,Borghate V B.Electromagnetic 79:19 compatibility estimator using scaled conjugate gradient back-prop- B]ZangTZ,Zhang L J,Zhang L,et al.Causes and countermeas- agation based artificial neural network.IEEE Trans Ind Inform, ures of casualty accident induced by unexpected factors in limited 2017,13(3):1036 job space.J Nanjing Univ Technol Nat Sci Ed,2015,27(3): [15]Akhtar N,Mian A.Threat of adversarial attacks on deep learning 103 in computer vision:a survey.IEEE Access,2018,6:14410 (臧铁柱,张礼敬,张丽,等.有限空间作业意外伤亡事故的 06]Razzak M I,Naz S,Zaib A.Deep learning for medical image 成因及其对策.南京工业大学学报(自然科学版),2005,27 processing:overview,challenges and the future /Classification (3):103) in Bioapps:Automation of Decision Making.Springer,2018 [3]Sun Y J,Shen HG.Industrial Ventilation.4th Ed.Beijing:Chi- [17]Wang H,Zhao Y,Xu Y M,et al.Cross-anguage speech attrib- na Architecture Building Press,2010 ute detection and phone recognition for Tibetan using deep learn- (孙一坚,沈恒根.工业通风.4版.北京:中国建筑工业出版 ing /The 9th International Symposium on Chinese Spoken Lan- 社,2010) guage Processing.Singapore,2014:474工程科学学报,第 41 卷,第 6 期 图 8 误差直方图 Fig. 8 Error histogram 进行分类,相较于其他的机器学习方法,深度学习无 需人工进行特征提取且学习特征更加全面,保证了 识别的准确性. ( 2) 通过多次试验得出深层神经网络在具有 3 个隐藏层,且每层有 9 个神经元的情况下分类效果 较好,且能保证较快的训练速度. ( 3) 通过混淆矩阵分析和 ROC 曲线分析可知 深层神经网络对不同氧气浓度的 PPG 信号分类性 能优良,在训练集中分类准确率可达 97. 9% ,在验 证集中分类准确率可达 94. 8% ,在测试集中的分类 准确 率 为 92. 8% . 在全部的数据中准确率为 96. 3% ,认为能够准确的识别人体是否处于氧气耐 受极限状态. ( 4) 所搭建的深层神经网络进行低氧状态评估 只需采集 2000 个离散点( 4 s 采集完成) ,因而能够 保证整个评估过程在 4 s 内完成,从而使得作业人员 暴露在低氧环境中的时间减少,保证了作业人员的 安全. 参 考 文 献 [1] Burlet-Vienney D,Chinniah Y,Bahloul A,et al. Occupational safety during interventions in confined spaces. Safety Sci,2015, 79: 19 [2] Zang T Z,Zhang L J,Zhang L,et al. Causes and countermeas￾ures of casualty accident induced by unexpected factors in limited job space. J Nanjing Univ Technol Nat Sci Ed,2015,27 ( 3) : 103 ( 臧铁柱,张礼敬,张丽,等. 有限空间作业意外伤亡事故的 成因及其对策. 南京工业大学学报( 自然科学版) ,2005,27 ( 3) : 103) [3] Sun Y J,Shen H G. Industrial Ventilation. 4th Ed. Beijing: Chi￾na Architecture & Building Press,2010 ( 孙一坚,沈恒根. 工业通风. 4 版. 北京: 中国建筑工业出版 社,2010) [4] Zhou X F,Yang F,Guo L,et al. Analysis and preventive recom￾mendation of national confined space accidents due to asphyxiation and poisoning from 2014 to 2015. J Eniviron Occup Med,2018, 35( 8) : 735 ( 周兴藩,杨凤,郭玲,等. 2014—2015 年全国有限空间作业 中毒与窒息事故分析及预防建议. 环境与职业医学,2018,35 ( 8) : 735) [5] Mejías C,Jiménez D,Muoz A,et al. Clinical response of 20 people in a mining refuge: study and analysis of functional parame￾ters. Safety Sci,2014,63: 204 [6] Selman J,Spickett J,Jansz J,et al. An investigation into the rate and mechanism of incident of work-related confined space fatali￾ties. Safety Sci,2018,109: 333 [7] Li G J. Research on Human Heat Tolerance under Extreme Hot, Humid and Low-oxygen Environment [Dissertaion]. Tianjin: Tianjin University,2008 ( 李国建. 高温高湿低氧环境下人体热耐受性研究[学位论 文]. 天津: 天津大学,2008) [8] Kamshilin A A,Nippolainen E,Sidorov I S,et al. A new look at the essence of the imaging photoplethysmography. Sci Rep,2015, 5: 10494 [9] Shin H,Min S D. Feasibility study for the non-invasive blood pressure estimation based on ppg morphology: normotensive sub￾ject study. Biomed Eng Online,2017,16( 1) : 10 [10] Zhou Q,Tu H,Zuo J X. Blood oxygen saturation real-time moni￾toring system research based on PPG. Inform Res,2017,43( 3) : 75 ( 张强,涂浩,左佳鑫. 基于 PPG 的血氧饱和度实时监测系 统研究. 信息化研究,2017,43( 3) : 75) [11] Njoum H,Kyriacou P A. Photoplethysmography for the assess￾ment of haemorheology. Sci Rep,2017,7( 1) : 1406 [12] Ma J L,Wang C,Li Z J,et al. Study of measuring heart rate and respiration rate based on PPG. Opt Tech,2011,37( 3) : 309 ( 马俊领,王成,李章俊,等. 基于 PPG 的心率和呼吸频率 的测量研究. 光学技术,2011,37( 3) : 309) [13] Tamura T,Maeda Y,Sekine M,et al. Wearable photoplethys￾mographic sensors-past and present. Electronics,2014,3 ( 2) : 282 [14] Khadse C B,Chaudhari M A,Borghate V B. Electromagnetic compatibility estimator using scaled conjugate gradient back-prop￾agation based artificial neural network. IEEE Trans Ind Inform, 2017,13( 3) : 1036 [15] Akhtar N,Mian A. Threat of adversarial attacks on deep learning in computer vision: a survey. IEEE Access,2018,6: 14410 [16] Razzak M I,Naz S,Zaib A. Deep learning for medical image processing: overview,challenges and the future / / Classification in Bioapps: Automation of Decision Making. Springer,2018 [17] Wang H,Zhao Y,Xu Y M,et al. Cross-language speech attrib￾ute detection and phone recognition for Tibetan using deep learn￾ing / / The 9th International Symposium on Chinese Spoken Lan￾guage Processing. Singapore,2014: 474 · 228 ·
<<向上翻页向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有