In this paper,a method based on the theory of geomorphic development and drainage fractal dimension is proposed for testing the rationality of four threshold determination methods for catchment area:drainage density m...In this paper,a method based on the theory of geomorphic development and drainage fractal dimension is proposed for testing the rationality of four threshold determination methods for catchment area:drainage density method,channel branching method,minimum error in river source distance method,trial and error method.The results show that drainage density method and channel branch method have a good applicability in the watershed of the loess hilly landform and can reflect the accurate information of gully channel.River source minimum error method is controlled by the topography of the river source,which is only applicable to small watershed with little change of topography.Trial and error method is controlled by artificial vision and scale,so it is not suitable for the area with large changes of topography or the complex river network.展开更多
The purpose of this study was to investigate the effects of a traditional Japanese medicine Goshajinkigan (TJ-107) and Tokishigyakukagoshuyushokyoto (TJ-38) on warm sense threshold, cold sense threshold and the periph...The purpose of this study was to investigate the effects of a traditional Japanese medicine Goshajinkigan (TJ-107) and Tokishigyakukagoshuyushokyoto (TJ-38) on warm sense threshold, cold sense threshold and the peripheral blood flow. 31 healthy volunteers (control group: 9people, TJ-107 group: 12 people, TJ-38group:10 people) were examined. Drugs administered 2.5 g a dose. Analysis was before and after 1 hour dosage. The warm and cold sense threshold in the thenar of the non-handedness site of these subjects was measured using a thermostimulator (Intercross-200, Intercross Co., Tokyo, Japan). The peripheral blood flow in the finger of the non-handedness site of these subjects was measured using a full-field laser perfusion imager (FLPI, Moor Instruments Ltd., England). Control: The vehicle had no significant effect on the warm sense threshold, cold sense threshold and the peripheral blood flow. TJ-107: The warm sense threshold and cold sense threshold were significantly decreased, and the reaction latency of cold sense was significantly shortening. The peripheral blood flow was significantly increased second and third finger at 115.6%, 119.3%, respectively. TJ-38: The cold sense threshold and the reaction latency of cold sense were significantly increased. The peripheral blood flow was significantly increased second and third finger with 114.3%, 112.8%, respectively. These results suggest that TJ-107 and TJ-38 have effects on the changed warm sense threshold, cold sense threshold and increased peripheral blood flow.展开更多
目的针对数据中心网络(Data Center Network,DCN)中数据流量多导致大象流与老鼠流识别精确度低的问题,提出一种基于软件定义网络(Software Defined Networking,SDN)下两阶段大象流识别算法。方法将SDN与DCN结合,第一阶段,采用高斯分布...目的针对数据中心网络(Data Center Network,DCN)中数据流量多导致大象流与老鼠流识别精确度低的问题,提出一种基于软件定义网络(Software Defined Networking,SDN)下两阶段大象流识别算法。方法将SDN与DCN结合,第一阶段,采用高斯分布动态阈值优化算法,通过对数据包阈值的设定,计算大象流误检率与漏检率,不断优化得到最优阈值,以此识别出可疑大象流;第二阶段,在依据流传输速率与流持续时间精确得到大象流的基础上,提出阈值约束、流量检测机制、Count计数器等三方面改进对大象流识别阈值下限的约束,将网络中大象流的数据量与流持续时间进行周期内阈值计算,提高大象流的识别精确度。结果实验结果表明:算法与已有相关算法相比,第一阶段可疑大象流平均字节数比网络流平均字节数多11.3%;不同阈值下的算法准确度提高1.7%,不同网络流量下的大象流平均检测时间降低至6 ms以内。结论软件定义网络下两阶段大象流识别算法在第一阶段具有较强的大象流识别能力,同时算法的精确度有所提高,大象流的平均检测时间降低,提高了网络质量,能为进行网络流量调度策略的进一步研究提供相关性条件。展开更多
为加强河流生态流量阈值及保障率计算对天然水文情势变化的适应性,提出了基于水文集合分位数的生态流量计算方法。根据河流长序列实测径流数据,在年际和年内划分不同水文情景,基于生态流量水文集合的中位数和下四分位数确定年内汛期、...为加强河流生态流量阈值及保障率计算对天然水文情势变化的适应性,提出了基于水文集合分位数的生态流量计算方法。根据河流长序列实测径流数据,在年际和年内划分不同水文情景,基于生态流量水文集合的中位数和下四分位数确定年内汛期、非汛期生态流量阈值下限,以丰、平水年组的计算结果分别作为平、枯水年组的生态流量理想值,结合生态流量中长期和短期保障率,计算分析河流在不同时间尺度下的生态流量保障程度。抚河下游应用结果表明:综合水情变化的生态流量阈值和多时间尺度保障率分析方法能够更好地适应天然河流年际和年内水情变化,更准确识别河流水文健康状况;丰、平、枯水年组抚河下游年均生态流量阈值下限分别为156.49、90.43、29.54 m 3/s;不同时间尺度下抚河下游生态流量的保障程度均较低,中长期保障率大于短期保障率,中长期和短期保障率分别为63.64%~93.11%、42.50%~91.85%。展开更多
[Objective] The aim was to al eviate the constructed wetland clogging problems and to explore to the effects of hydraulic loading on wetland clogging. [Method] The experiment, through artificial soil columns, simulate...[Objective] The aim was to al eviate the constructed wetland clogging problems and to explore to the effects of hydraulic loading on wetland clogging. [Method] The experiment, through artificial soil columns, simulated vertical flow arti-ficial wetland, set four hydraulic load level at 50, 100, 150 and 200 cm/d, to identify the impact of hydraulic loading on wetland clogging and to explore the factors run-ning threshold. [Result] The results show that the different levels of hydraulic loading have greater impact; in the constructed wetland clogging process under high hy-draulic loading of 200 cm/d, the effective life was only six months, and the single factor can be speculated that the threshold of the hydraulic load should be at 100-150 cm/d; system can last for six months at low hydraulic loading and last for three months at medium hydraulic load. [Conclusion] The research provides references for wetland clogging experiments in future.展开更多
针对同步定位与地图建立(simultaneous localization and mapping,SLAM)算法在动态环境下存在位姿估计和地图构建误差较大的问题,提出一种光流语义分割方法用于增加动态场景下的可运行性。将ORB-SLAM2系统与YOLOv5模型结合,对传入图像...针对同步定位与地图建立(simultaneous localization and mapping,SLAM)算法在动态环境下存在位姿估计和地图构建误差较大的问题,提出一种光流语义分割方法用于增加动态场景下的可运行性。将ORB-SLAM2系统与YOLOv5模型结合,对传入图像提取特征点的同时将YOLOv5网络模型语义分割后的物体分为高、中、低动态物体。利用运动一致性检测算法,对三种检测物体动态阈值判断,辨别其是否需要剔除特征点,增加ORB-SLAM2算法在动态环境下的运行精度。为加快系统运行速度,用LK光流法加快普通帧与普通帧之间的匹配,其原理为使用LK光流匹配特征点代替ORB特征点匹配,大大的缩小运行时间,同时运行误差变化不大。实验在TUM数据集下测试,平均每一帧提取2000个特征点,在增加LK光流后缩短0.01 s以上,若在900帧数据集下,可缩短9 s.其绝对轨迹误差对比于ORB-SLAM2和DS-SLAM平均提升在95%与30%以上,证明了算法在动态场景下良好的运行精度与鲁棒性。展开更多
基金Supported by Doctoral and Youth Science and Technology Fund Project in 2021(2021K007).
文摘In this paper,a method based on the theory of geomorphic development and drainage fractal dimension is proposed for testing the rationality of four threshold determination methods for catchment area:drainage density method,channel branching method,minimum error in river source distance method,trial and error method.The results show that drainage density method and channel branch method have a good applicability in the watershed of the loess hilly landform and can reflect the accurate information of gully channel.River source minimum error method is controlled by the topography of the river source,which is only applicable to small watershed with little change of topography.Trial and error method is controlled by artificial vision and scale,so it is not suitable for the area with large changes of topography or the complex river network.
文摘The purpose of this study was to investigate the effects of a traditional Japanese medicine Goshajinkigan (TJ-107) and Tokishigyakukagoshuyushokyoto (TJ-38) on warm sense threshold, cold sense threshold and the peripheral blood flow. 31 healthy volunteers (control group: 9people, TJ-107 group: 12 people, TJ-38group:10 people) were examined. Drugs administered 2.5 g a dose. Analysis was before and after 1 hour dosage. The warm and cold sense threshold in the thenar of the non-handedness site of these subjects was measured using a thermostimulator (Intercross-200, Intercross Co., Tokyo, Japan). The peripheral blood flow in the finger of the non-handedness site of these subjects was measured using a full-field laser perfusion imager (FLPI, Moor Instruments Ltd., England). Control: The vehicle had no significant effect on the warm sense threshold, cold sense threshold and the peripheral blood flow. TJ-107: The warm sense threshold and cold sense threshold were significantly decreased, and the reaction latency of cold sense was significantly shortening. The peripheral blood flow was significantly increased second and third finger at 115.6%, 119.3%, respectively. TJ-38: The cold sense threshold and the reaction latency of cold sense were significantly increased. The peripheral blood flow was significantly increased second and third finger with 114.3%, 112.8%, respectively. These results suggest that TJ-107 and TJ-38 have effects on the changed warm sense threshold, cold sense threshold and increased peripheral blood flow.
文摘目的针对数据中心网络(Data Center Network,DCN)中数据流量多导致大象流与老鼠流识别精确度低的问题,提出一种基于软件定义网络(Software Defined Networking,SDN)下两阶段大象流识别算法。方法将SDN与DCN结合,第一阶段,采用高斯分布动态阈值优化算法,通过对数据包阈值的设定,计算大象流误检率与漏检率,不断优化得到最优阈值,以此识别出可疑大象流;第二阶段,在依据流传输速率与流持续时间精确得到大象流的基础上,提出阈值约束、流量检测机制、Count计数器等三方面改进对大象流识别阈值下限的约束,将网络中大象流的数据量与流持续时间进行周期内阈值计算,提高大象流的识别精确度。结果实验结果表明:算法与已有相关算法相比,第一阶段可疑大象流平均字节数比网络流平均字节数多11.3%;不同阈值下的算法准确度提高1.7%,不同网络流量下的大象流平均检测时间降低至6 ms以内。结论软件定义网络下两阶段大象流识别算法在第一阶段具有较强的大象流识别能力,同时算法的精确度有所提高,大象流的平均检测时间降低,提高了网络质量,能为进行网络流量调度策略的进一步研究提供相关性条件。
文摘为加强河流生态流量阈值及保障率计算对天然水文情势变化的适应性,提出了基于水文集合分位数的生态流量计算方法。根据河流长序列实测径流数据,在年际和年内划分不同水文情景,基于生态流量水文集合的中位数和下四分位数确定年内汛期、非汛期生态流量阈值下限,以丰、平水年组的计算结果分别作为平、枯水年组的生态流量理想值,结合生态流量中长期和短期保障率,计算分析河流在不同时间尺度下的生态流量保障程度。抚河下游应用结果表明:综合水情变化的生态流量阈值和多时间尺度保障率分析方法能够更好地适应天然河流年际和年内水情变化,更准确识别河流水文健康状况;丰、平、枯水年组抚河下游年均生态流量阈值下限分别为156.49、90.43、29.54 m 3/s;不同时间尺度下抚河下游生态流量的保障程度均较低,中长期保障率大于短期保障率,中长期和短期保障率分别为63.64%~93.11%、42.50%~91.85%。
基金Supported by National Natural Science Foundation of China(41071214)~~
文摘[Objective] The aim was to al eviate the constructed wetland clogging problems and to explore to the effects of hydraulic loading on wetland clogging. [Method] The experiment, through artificial soil columns, simulated vertical flow arti-ficial wetland, set four hydraulic load level at 50, 100, 150 and 200 cm/d, to identify the impact of hydraulic loading on wetland clogging and to explore the factors run-ning threshold. [Result] The results show that the different levels of hydraulic loading have greater impact; in the constructed wetland clogging process under high hy-draulic loading of 200 cm/d, the effective life was only six months, and the single factor can be speculated that the threshold of the hydraulic load should be at 100-150 cm/d; system can last for six months at low hydraulic loading and last for three months at medium hydraulic load. [Conclusion] The research provides references for wetland clogging experiments in future.
