Projection of the East Asian westerly jet under six global warming targets

东亚高空急流(简称急流)对夏季东亚地区的天气和气候有着重要影响。本文利用CMIP5的历史气候模拟试验和RCP8.5路径下的未来气候变化预估试验数据,预估了急流在6个全球变暖阈值(1.5℃,2.0℃,2.5℃,3.0℃,3.5℃和4.0℃)下相对于当代气候的变化情况。结果表明东亚高空西风在1.5℃阈值下略微减弱。在2.0℃阈值下,西风在急流轴(约40°N)南侧增强,北侧减弱。这种变化趋势在2.5℃和更高的变暖阈值下愈加明显,使急流轴逐渐向南移动,但急流强度变化不大。研究表明,在急流的入口和出口区,对流层中上层大气升温速度相对较慢,导致在急流轴南(北)侧出现向南(北)的负(正)经向温度梯度,使西风在南(北)侧增强(减弱).

Projected changes in the western North Pacific subtropical high under six global warming targets

西北太平洋副热带高压(简称西太副高)是影响东亚夏季天气气候的关键环流系统。本文利用CMIP5的历史气候模拟试验和RCP8.5路径下的未来气候变化预估试验数据,采用扰动位势高度,流函数等多种变量描述西太副高,分析了西太副高在6个全球变暖阈值(1.5℃, 2.0℃, 2.5℃,3.0℃, 3.5℃和4.0℃)下相对于当代气候的变化情况。在对流层中层(500 h Pa),西太副高在1.5℃阈值下几乎没有变化,而在2.0℃阈值下迅速减弱并东退约2.5°。当升温大于2.5℃时,西太副高呈线性减弱趋势,在4.0℃阈值下将东退约6.0°。在对流层低层(850 h Pa),西太副高在1.5℃阈值下增强西伸,但在升温到2.0℃的过程中变化不大。当变暖达到4.0℃阈值时,西太副高将西伸约2.0°。

Projected changes in Köppen-Trewartha climate zones under 1.5-4℃global warming targets over mid-high latitudes of Northern Asia using an ensemble of RegCM4 simulations

Mid-high latitude Northern Asia is one of the most vulnerable and sensitive areas to global warming,but relatively less studied previously.We used an ensemble of a regional climate model(RegCM4)projections to assess future changes in surface air temperature,precipitation and Köppen-Trewartha(K-T)climate types in Northern Asia under the 1.5-4℃global warming targets.RegCM4 is driven by five CMIP5 global models over an East Asia domain at a grid spacing of 25 km.Validation of the present day(1986-2005)simulations shows that the ensembles of RegCM4(ensR)and driving GCMs(ensG)reproduce the major characters of the observed temperature,precipitation and K-T climate zones reasonably well.Greater and more realistic spatial detail is found in RegCM4 compared to the driving GCMs.A general warming and overall increases in precipitation are projected over the region,with these changes being more pronounced at higher warming levels.The projected warming by ensR shows different spatial patterns,and is in general lower,compared to ensG in most months of the year,while the percentage increases of precipitation are maximum during the cold months.The future changes in K-T climate zones are characterized by a substantial expansion of Dc(temperature oceanic)and retreat of Ec(sub-arctic continental)over the region,reaching∼20%under the 4℃warming level.The most notable change in climate types in ensR is found over Japan(∼60%),followed by Southern Siberia,Mongolia,and the Korean Peninsula(∼40%).The largest change in the K-T climate types is found when increasing from 2 to 3℃.The results will help to better assess the impacts of climate change and in implementation of appropriate adaptation measures over the region.

Long-term acupuncture treatment has a multitargeting regulation on multiple brain regions in rats with Alzheimer＇s disease：a positron emission tomography study

The acute effect of acupuncture on Alzheimer＇s disease,i.e.,on brain activation during treatment,has been reported.However,the effect of long-term acupuncture on brain activation in Alzheimer＇s disease is unclear.Therefore,in this study,we performed long-term needling at Zusanli（ST36）or a sham point（1.5 mm lateral to ST36）in a rat Alzheimer＇s disease model,for 30 minutes,once per day,for 30 days.The rats underwent 18F-fluorodeoxyglucose positron emission tomography scanning.Positron emission tomography images were processed with SPM2.The brain areas activated after needling at ST36 included the left hippocampus,the left orbital cortex,the left infralimbic cortex,the left olfactory cortex,the left cerebellum and the left pons.In the sham-point group,the activated regions were similar to those in the ST36 group.However,the ST36 group showed greater activation in the cerebellum and pons than the sham-point group.These findings suggest that long-term acupuncture treatment has targeted regulatory effects on multiple brain regions in rats with Alzheimer＇s disease.

Asian climate change under 1.5-4 ℃ warming targets

在三种 RCP 情形下面基于 18 个 CMIP5 模型的模拟，这篇文章在全球温暖的上下文在亚洲上在吝啬的温度和降水和他们的极端调查变化 1.5-4 的目标吗？

Research on Motion Attention Fusion Model-Based Video Target Detection and Extraction of Global Motion Scene

For target detection algorithm under global motion scene, this paper suggests a target detection algorithm based on motion attention fusion model. Firstly, the motion vector field is pre-processed by accumulation and median filter;Then, according to the temporal and spatial character of motion vector, the attention fusion model is defined, which is used to detect moving target;Lastly, the edge of video moving target is made exactly by morphologic operation and edge tracking algorithm. The experimental results of different global motion video sequences show the proposed algorithm has a better veracity and speedup than other algorithm.

Long-Term Projections of Global Food Requirements: Why Were We Wrong?

This article reports on and analyzes long-term projections of world food requirements compared with observed 2000 data reported by the United Nations’ Food and Agriculture Organization. The importance of this “post-mortem” is to strengthen the case for carrying out long-term projections of essential resources—food, energy, and non-fuel minerals— because of the long-lead times needed to insure that adequate global output levels of these basic ingredients of living standards will be met. This study should prove useful to those preparing today’s long-term projections, with world population projected to rise to over 9bn by mid-century.

THE NINTH FIVE-YEAR PLAN AND LONG-TERM TARGET BY 2010 FOR CHINA'S FOOD INDUSTRY

The"Ninth Five-Year Plan for theNational Economy and SocialDevelopment and the Long-TermTarget by the Year 2010" adopted at theFourth Session of the Eighth National People’sCongress have opened up a vast new worldfor the development of China’s food industry.The food industry should seize the opportunityto strengthen itself and speed up itsdevelopment for meeting the needs ofimproving the people’s living standards andperform well in modernized socialistconstruction.

A spawning particle filter for defocused moving target detection in GNSS-based passive radar

Global Navigation Satellite System(GNSS)-based passive radar(GBPR)has been widely used in remote sensing applications.However,for moving target detection(MTD),the quadratic phase error(QPE)introduced by the non-cooperative target motion is usually difficult to be compensated,as the low power level of the GBPR echo signal renders the estimation of the Doppler rate less effective.Consequently,the moving target in GBPR image is usually defocused,which aggravates the difficulty of target detection even further.In this paper,a spawning particle filter(SPF)is proposed for defocused MTD.Firstly,the measurement model and the likelihood ratio function(LRF)of the defocused point-like target image are deduced.Then,a spawning particle set is generated for subsequent target detection,with reference to traditional particles in particle filter(PF)as their parent.After that,based on the PF estimator,the SPF algorithm and its sequential Monte Carlo(SMC)implementation are proposed with a novel amplitude estimation method to decrease the target state dimension.Finally,the effectiveness of the proposed SPF is demonstrated by numerical simulations and pre-liminary experimental results,showing that the target range and Doppler can be estimated accurately.

融合目标检测与DWA算法的AGV路径

针对AGV(Automated Guided Vehicle)叉车处于环境信息未知或环境动态变化情况下的路径规划及导航问题,文中提出了一种由YOLOv5(You Only Look Once version 5)目标检测算法获取目标位置。根据目标位置规划出全局基础路径,再融合DWA(Dynamic Window Approach)局部动态路径规划算法进行AGV路径规划与导航,使AGV叉车在未知环境或局部环境信息未知的环境中能快速识别目标位置并完成路径规划到达目标位置。实验结果表明,相较于改进前方法,文中所提方法在路径长度、耗费时间以及AGV叉车航向误差方面均有良好表现,路径长度平均减少12%,耗费时间平均减少约5%且AGV航向与目标航向的平均误差在5°以内。所提方法提高了AGV叉车在未知环境中的工作效率以及工作灵活性。

基于双曲嵌入的露天矿区暗光环境下道路多目标检测模型

露天矿环境特殊,道路场景复杂多变,在光照不足时会导致矿区道路多目标识别不清、定位不准,进而影响检测效果,给矿区无人矿用卡车的安全行驶带来严重安全隐患。目前的道路障碍物检测模型不能有效解决矿区暗光环境对模型检测效果的影响,同时对矿区小目标障碍物的识别也有较大误差,不适用于矿区特殊环境下障碍物的检测与识别。针对上述问题,提出了一种基于双曲嵌入的露天矿区暗光环境下多目标检测模型。首先,在模型的图像预处理阶段引入卷积神经网路Retinex-Net对暗图像进行增强,提高图像清晰度;然后,针对数据集中特征过多而无重点偏好的问题,在加强特征提取部分添加全局注意力机制,聚集3个维度上更关键的特征信息;最后,在检测模型预测阶段引入双曲全连接层,以减少特征丢失,并防止过拟合现象。实验结果表明:(1)基于双曲嵌入的露天矿区暗光环境下道路多目标检测模型不仅对露天矿区暗光环境下的大尺度目标具有较高的分类与定位精度,对矿用卡车及较远距离的小尺度目标即行人也可准确检测与定位,能够满足无人矿用卡车在矿区特殊环境下驾驶的安全需求。(2)模型的检测准确率达98.6%,检测速度为51.52帧/s,较SSD、YOLOv4、YOLOv5、YOLOx、YOLOv7分别提高20.31%,18.51%,10.53%,8.39%,13.24%,对于矿区道路上的行人、矿用卡车及挖机的检测精度达97%以上。

随机Lur’e时滞网络的指数同步与脉冲控制

本文研究了一类具有不对称耦合和时变时滞随机Lur’e动态网络的全局指数同步问题.为了使网络中的所有Lur’e系统达到同步,文章提出了一种实用的脉冲控制策略,利用平均脉冲间隔概念、Perron-Frobenius定理、比较原理、参数变分法和无参照系统策略,给出了保证网络实现全局指数同步的充分条件.基于脉冲信号既能促进网络群聚行为同时也可能阻碍网络同步这一事实,本文分别讨论了积极脉冲效应和消极脉冲效应作用下网络实现同步的判定方法,同时给出了不同脉冲效应作用下网络的收敛速度.提供了数值仿真证明主要结果和控制策略的有效性.

基于改进YOLOv5的船舶多尺度SAR图像检测算法

针对复杂场景下合成孔径雷达(Synthetic aperture radar, SAR)图像船舶目标像素尺度差异大和船舶密集排列造成目标漏检的问题,提出一种基于改进YOLOv5的船舶多尺度SAR图像检测算法。对于YOLOv5的颈部网络,采用双向特征金字塔结构(Bi-directional feature pyramid network, BiFPN)提升网络多尺度特征融合能力,并在其自下而上的特征融合支路中,基于深度可分离卷积(Depthwise separable convolution, DSC)和通道MLP构建EC-MLP(Enhanced channel-MLP)模块,从而丰富语义信息,提供更充分的船舶目标上下文特征;引入全局注意力机制(Global attention mechanism, GAM),使网络对输入特征进行针对性提取并运算,减少网络的信息丢失;此外,使用SIoU损失函数进一步提高网络的训练收敛速度和检测精度。在SSDD和HRSID数据集上与其他8种方法(Faster R-CNN、Libra R-CNN、FCOS、YOLOv5s、PP-YOLOv2、YOLOX-s、PP-YOLOE-s和YOLOv7-tiny)进行对比实验。实验结果表明:改进后算法在SSDD数据集上的AP50达到了96.7%,在HRSID数据集上AP50达到了95.6%,优于对比方法。

基于双目视觉的六边形芯片晶圆台高度差测量

针对检测中晶圆台可能影响检测精度的问题,提出了一种基于双目视觉测量芯片晶圆台的测量方法。由于没有现成的六边形芯片数据集,因此采集的样本主要来自由CAD软件绘制生成的理想数据集,然后对采集的数据集进行训练,训练结果值符合检测识别的要求,且准确识别了六边形芯片。最后,利用YOLOv5s网络框架和全局立体匹配算法来对同一直线边缘芯片进行匹配测量,通过求差,计算是否含有高度差,由高度差可以判断出晶圆台是否在同一水平面上。多次实验的结果表明,此方法可以准确检测到六边形芯片晶圆台的高度差。

融合多层次浅层信息的航拍小目标检测

针对小目标检测及目标被遮挡的问题,本文基于VisDrone2019数据集构建相应交通场景,提出一种小目标检测算法.首先,充分利用主干网络的浅层特征改善小目标漏检的问题,通过在YOLOv7算法原有的网络结构上增加小目标检测层P2,并在P2小目标检测层的模型上为特征融合网络添加多层次浅层信息融合模块,从而提高算法小目标检测效果.其次,使用全局上下文模块构建目标与全局上下文的联系,增强模型区分目标与背景的能力,改善目标因遮挡而出现特征缺失情况下的被检测效果.最后,本文采用专为小目标设计的损失函数NWD代替基线模型中的CIoU损失函数,从而解决了IoU本身及其扩展对微小物体的位置偏差非常敏感的问题.实验表明,改进后的YOLOv7模型在航拍小目标数据集VisDrone2019(测试集和验证集)上面mAP.5:.95分别有2.3%和2.8%的提升,取得了十分优异的检测效果.

基于YOLOv4改进特征融合及全局感知的目标检测算法

YOLOv4算法在检测速度和精度上达到了很好的平衡,但仍存在着定位框不准确、检测率低的问题,尤其是在检测目标较小、尺度变化大的情况下。针对以上问题,提出一种新的基于YOLOv4改进的目标检测算法。该算法采用改进的特征融合模块(path aggregation network combined with bi-directional feature pyramid network,P-Bifpn)代替PANet(path aggregation network),增加跨尺度连接的同时在输出端引入权重,增强重要特征的表现力,解决由多尺度变化而引起的精度下降。然后,采用新的全局注意力机制(global association network,GANet),在减少平均池化与计算量的同时增强Sigmoid函数输出,加强模型对目标上下文关系的学习,减少噪声干扰和全局信息的损失。试验采用RSOD、NWPU VHR-10数据集,平均检测精度分别提升了约5%和3%;泛化试验采用VOC2007+2012公共数据集,平均检测精度提升了约0.6%。试验结果表明改进的算法能够有效提高模型的检测能力。

基于RANSAC和三维谱峰分析的全姿态散射中心建模

全姿态散射中心模型是一种性能优良的光学区复杂目标电磁散射参数化模型。针对传统的基于候选点筛选和聚类的全姿态散射中心建模方法易出现虚假散射中心和遗漏真实散射中心的问题,该文提出了一种基于目标三维空间电磁散射强度场谱峰分析的建模方法。首先,基于目标多视一维散射中心参数,利用随机采样一致性(RANSAC)方法和Parzen窗函数方法估计目标在三维空间中的电磁散射强度场。然后,通过谱峰分析、散射中心关联和多视量测融合,得到全姿态三维散射中心的位置。最后,利用二值形态学处理修正全姿态散射中心的角度可见性,估计全姿态散射中心的散射系数和类型参数。仿真结果表明,该文方法所提取的全姿态散射中心与目标几何结构具有极强的关联性,相较传统方法,在缩减三维散射中心数量的同时提升了模型的表示精度。

基于全局注意力机制的Robust-PointPillars三维目标检测

提出了一种基于全局注意力机制的Robust-PointPillars三维目标检测方法,在智能驾驶的应用中,提高了目标检测的精度和鲁棒性。PointPillars等神经网络通过使用点云柱表示点云,具有实现三维目标检测的潜力。首先介绍了空间和通道双重注意力模块,以增强有学习价值的点云特征,解决了PointPillars缺乏点云柱内部学习机制和特征提取不足的问题;挤压与激励网络(SENet)模块的引入,使PointPillars对特征信息的学习理解能力得到进一步提高;最终,对受到干扰或缺失的传感器信号进行抑制,并利用全局注意力算法来提高鲁棒性。基于KITTI数据集上的目标检测结果,本文算法具有良好的目标检测精度和鲁棒性。

Summed volume region selection based three-dimensional automatic target recognition for airborne LIDAR

Airborne LIDAR can flexibly obtain point cloud data with three-dimensional structural information,which can improve its effectiveness of automatic target recognition in the complex environment.Compared with 2D information,3D information performs better in separating objects and background.However,an aircraft platform can have a negative influence on LIDAR obtained data because of various flight attitudes,flight heights and atmospheric disturbances.A structure of global feature based 3D automatic target recognition method for airborne LIDAR is proposed,which is composed of offline phase and online phase.The performance of four global feature descriptors is compared.Considering the summed volume region(SVR) discrepancy in real objects,SVR selection is added into the pre-processing operations to eliminate mismatching clusters compared with the interested target.Highly reliable simulated data are obtained under various sensor’s altitudes,detection distances and atmospheric disturbances.The final experiments results show that the added step increases the recognition rate by above 2.4% and decreases the execution time by about 33%.