针对现有作物三维重建方法存在重建时间长、重建效果差的问题,构建了一种基于改进定向快速旋转描述-即时定位与地图构建二代系统(oriented FAST and rotated BRIEF-SLAM2,ORB-SLAM2)算法的苗期玉米植株重建方法。利用快速最近邻搜索库(f...针对现有作物三维重建方法存在重建时间长、重建效果差的问题,构建了一种基于改进定向快速旋转描述-即时定位与地图构建二代系统(oriented FAST and rotated BRIEF-SLAM2,ORB-SLAM2)算法的苗期玉米植株重建方法。利用快速最近邻搜索库(fast library for approximate nearest neighbors,FLANN)算法结合随机抽样一致性算法(random sample consensus,RANSAC)对苗期玉米植株图像进行特征匹配,并结合多视角立体算法(multiple view-tereo,MVS)实现对苗期玉米植株的稠密重建。在此基础上,利用重建模型对两个玉米品种植株的主要构型参数进行获取,并与实测值进行对比,以验证构型获取方法的准确性和有效性。试验结果表明:FLANN+RANSAC算法进行特征匹配的正确率是89.00%,点云平均稠密重建点数为7.13×105个,平均重建时间仅为15.32 min,利用三维重建模型进行苗期玉米植株构型参数提取的误差均能控制在10%以内,且与人工实测值具有较好的相关性。该算法重建时间短,且重建精度较高,能够为苗期玉米植株的构型获取提供理论依据和技术支撑。展开更多
针对传统视觉simultaneous localization and mapping(SLAM)回环检测算法在光照变化、动态场景及视角变化等复杂环境下容易出现定位精度下降和累积误差增大的问题,提出一种基于MobileNetV3的回环检测算法。利用预训练的MobileNetV3模型...针对传统视觉simultaneous localization and mapping(SLAM)回环检测算法在光照变化、动态场景及视角变化等复杂环境下容易出现定位精度下降和累积误差增大的问题,提出一种基于MobileNetV3的回环检测算法。利用预训练的MobileNetV3模型提取图像特征,并通过主成分分析(PCA)和白化处理降低特征向量维度,提升计算效率。采用余弦相似度计算图像特征之间的相似度矩阵,并根据设定阈值判断是否出现回环。在New College和City Centre数据集上的实验结果表明,基于MobileNetV3的回环检测算法在几种对比算法中表现最优,与基于视觉词袋模型(BOVW)的回环检测算法相比,所提算法在两个数据集上的检测准确率分别提高18.5%和19.3%,检测速度分别提高30.6%和34.4%,能更好满足视觉SLAM对准确性和实时性的要求。最后将此算法应用到oriented FAST and rotated brief SLAM2(ORB-SLAM2)中,替换其原有的基于视觉词袋模型的回环检测算法,并在EuRoC数据集上测试改进后的ORB-SLAM2算法,实验结果表明,改进后的ORB-SLAM2算法定位精度比原算法提升23.8%,生成的轨迹曲线更接近真实轨迹,验证了所提算法在SLAM系统中的可行性和有效性。展开更多
A switch from avian-typeα-2,3 to human-typeα-2,6 receptors is an essential element for the initiation of a pandemic from an avian influenza virus.Some H9N2 viruses exhibit a preference for binding to human-typeα-2,...A switch from avian-typeα-2,3 to human-typeα-2,6 receptors is an essential element for the initiation of a pandemic from an avian influenza virus.Some H9N2 viruses exhibit a preference for binding to human-typeα-2,6 receptors.This identifies their potential threat to public health.However,our understanding of the molecular basis for the switch of receptor preference is still limited.In this study,we employed the random forest algorithm to identify the potentially key amino acid sites within hemagglutinin(HA),which are associated with the receptor binding ability of H9N2 avian influenza virus(AIV).Subsequently,these sites were further verified by receptor binding assays.A total of 12 substitutions in the HA protein(N158D,N158S,A160 N,A160D,A160T,T163I,T163V,V190T,V190A,D193 N,D193G,and N231D)were predicted to prefer binding toα-2,6 receptors.Except for the V190T substitution,the other substitutions were demonstrated to display an affinity for preferential binding toα-2,6 receptors by receptor binding assays.Especially,the A160T substitution caused a significant upregulation of immune-response genes and an increased mortality rate in mice.Our findings provide novel insights into understanding the genetic basis of receptor preference of the H9N2 AIV.展开更多
To enhance the accuracy of nearshore data products obtained from nadir radar altimeters,we introduce a novel two-step retracking algorithm for reconstructed waveforms.This approach utilizes Empirical Mode Decompositio...To enhance the accuracy of nearshore data products obtained from nadir radar altimeters,we introduce a novel two-step retracking algorithm for reconstructed waveforms.This approach utilizes Empirical Mode Decomposition(EMD)to extract trend information from the trailing edge of the waveform.Reconstructed waveforms are formed by linking the leading and trailing edge trend information.The retracking process consists of two steps:the first step focuses on retracking a segment of the leading edge to obtain 4 crucial a priori parameters.In the second step,retracking incorporates both the leading and trailing edges using the previously acquired a priori information.We tested the algorithm using data from the HY-2B radar altimeter.Results indicate that the proposed two-step retracking algorithm outperforms the Maximum Likelihood Estimation(MLE4)algorithm currently used in the operational processing of the HY-2B radar altimeter,as well as the Adaptive Leading Edge Subwaveform(ALES)algorithm,in terms of significant wave height(SWH)and sea level anomalies(SLA).Specifically,the standard deviation of the difference in SWH is reduced by 14%,and the standard deviation of the difference in SLA is reduced by approximately 18%.The two-step retracking algorithm effectively leverages trailing edge information,reduces the influence of peak noise on the leading edge,and improves both the utilization and accuracy of the waveform retracking.展开更多
To solve the problems of deformation,micro-cracks,and residual tensile stress in laser cladding coatings,the technique of laser cladding with Fe-based memory alloy can be considered.However,the process of in-situ synt...To solve the problems of deformation,micro-cracks,and residual tensile stress in laser cladding coatings,the technique of laser cladding with Fe-based memory alloy can be considered.However,the process of in-situ synthesis of Fe-based memory alloy coatings is extremely complex.At present,there is no clear guidance scheme for its preparation process,which limits its promotion and application to some extent.Therefore,in this study,response surface methodology(RSM)was used to model the response surface between the target values and the cladding process parameters.The NSGA-2 algorithm was employed to optimize the process parameters.The results indicate that the composite optimization method consisting of RSM and the NSGA-2 algorithm can establish a more accurate model,with an error of less than 4.5%between the predicted and actual values.Based on this established model,the optimal scheme for process parameters corresponding to different target results can be rapidly obtained.The prepared coating exhibits a uniform structure,with no defects such as pores,cracks,and deformation.The surface roughness and microhardness of the coating are enhanced,the shaping quality of the coating is effectively improved,and the electrochemical corrosion performance of the coating in 3.5%NaCl solution is obviously better than that of the substrate,providing an important guide for engineering applications.展开更多
针对燃煤电厂参与调峰负荷波动较大,出口SO2浓度控制效果不佳的问题,建立了一种基于捕鱼优化算法(catch fish optimization algorithm,CFOA)优化融合神经网络的出口SO2浓度预测模型。首先使用互信息算法筛选由机理分析得到的特征...针对燃煤电厂参与调峰负荷波动较大,出口SO2浓度控制效果不佳的问题,建立了一种基于捕鱼优化算法(catch fish optimization algorithm,CFOA)优化融合神经网络的出口SO2浓度预测模型。首先使用互信息算法筛选由机理分析得到的特征变量,并通过逐次变分模态分解对筛选后的辅助变量进行分解重构,保留相关性较大的重构分量作为输入变量。随后采用双向时间卷积网络、双向门控循环单元与多头自注意力机制构建融合神经网络模型,通过CFOA对模型超参数寻优以进一步提高精度。最后使用某660 MW燃煤电厂历史运行数据进行对比实验,实验结果表明,该模型在出口SO2浓度剧烈波动的工况下仍能实现较好的预测效果。同多种模型对比,该模型具有更小的误差和更高的预测精度,体现出其在复杂变化环境中的鲁棒性和可靠性。展开更多
摘要针对现有作物三维重建方法存在重建时间长、重建效果差的问题,构建了一种基于改进定向快速旋转描述-即时定位与地图构建二代系统(oriented FAST and rotated BRIEF-SLAM2,ORB-SLAM2)算法的苗期玉米植株重建方法。利用快速最近邻搜索库(fast library for approximate nearest neighbors,FLANN)算法结合随机抽样一致性算法(random sample consensus,RANSAC)对苗期玉米植株图像进行特征匹配,并结合多视角立体算法(multiple view-tereo,MVS)实现对苗期玉米植株的稠密重建。在此基础上,利用重建模型对两个玉米品种植株的主要构型参数进行获取,并与实测值进行对比,以验证构型获取方法的准确性和有效性。试验结果表明:FLANN+RANSAC算法进行特征匹配的正确率是89.00%,点云平均稠密重建点数为7.13×105个,平均重建时间仅为15.32 min,利用三维重建模型进行苗期玉米植株构型参数提取的误差均能控制在10%以内,且与人工实测值具有较好的相关性。该算法重建时间短,且重建精度较高,能够为苗期玉米植株的构型获取提供理论依据和技术支撑。
摘要针对传统视觉simultaneous localization and mapping(SLAM)回环检测算法在光照变化、动态场景及视角变化等复杂环境下容易出现定位精度下降和累积误差增大的问题,提出一种基于MobileNetV3的回环检测算法。利用预训练的MobileNetV3模型提取图像特征,并通过主成分分析(PCA)和白化处理降低特征向量维度,提升计算效率。采用余弦相似度计算图像特征之间的相似度矩阵,并根据设定阈值判断是否出现回环。在New College和City Centre数据集上的实验结果表明,基于MobileNetV3的回环检测算法在几种对比算法中表现最优,与基于视觉词袋模型(BOVW)的回环检测算法相比,所提算法在两个数据集上的检测准确率分别提高18.5%和19.3%,检测速度分别提高30.6%和34.4%,能更好满足视觉SLAM对准确性和实时性的要求。最后将此算法应用到oriented FAST and rotated brief SLAM2(ORB-SLAM2)中,替换其原有的基于视觉词袋模型的回环检测算法,并在EuRoC数据集上测试改进后的ORB-SLAM2算法,实验结果表明,改进后的ORB-SLAM2算法定位精度比原算法提升23.8%,生成的轨迹曲线更接近真实轨迹,验证了所提算法在SLAM系统中的可行性和有效性。
基金supported by the National Natural Science Foundation of China(32273037 and 32102636)the Guangdong Major Project of Basic and Applied Basic Research(2020B0301030007)+4 种基金Laboratory of Lingnan Modern Agriculture Project(NT2021007)the Guangdong Science and Technology Innovation Leading Talent Program(2019TX05N098)the 111 Center(D20008)the double first-class discipline promotion project(2023B10564003)the Department of Education of Guangdong Province(2019KZDXM004 and 2019KCXTD001).
摘要A switch from avian-typeα-2,3 to human-typeα-2,6 receptors is an essential element for the initiation of a pandemic from an avian influenza virus.Some H9N2 viruses exhibit a preference for binding to human-typeα-2,6 receptors.This identifies their potential threat to public health.However,our understanding of the molecular basis for the switch of receptor preference is still limited.In this study,we employed the random forest algorithm to identify the potentially key amino acid sites within hemagglutinin(HA),which are associated with the receptor binding ability of H9N2 avian influenza virus(AIV).Subsequently,these sites were further verified by receptor binding assays.A total of 12 substitutions in the HA protein(N158D,N158S,A160 N,A160D,A160T,T163I,T163V,V190T,V190A,D193 N,D193G,and N231D)were predicted to prefer binding toα-2,6 receptors.Except for the V190T substitution,the other substitutions were demonstrated to display an affinity for preferential binding toα-2,6 receptors by receptor binding assays.Especially,the A160T substitution caused a significant upregulation of immune-response genes and an increased mortality rate in mice.Our findings provide novel insights into understanding the genetic basis of receptor preference of the H9N2 AIV.
基金The National Natural Science Foundation of China under contract No.42192531。
摘要To enhance the accuracy of nearshore data products obtained from nadir radar altimeters,we introduce a novel two-step retracking algorithm for reconstructed waveforms.This approach utilizes Empirical Mode Decomposition(EMD)to extract trend information from the trailing edge of the waveform.Reconstructed waveforms are formed by linking the leading and trailing edge trend information.The retracking process consists of two steps:the first step focuses on retracking a segment of the leading edge to obtain 4 crucial a priori parameters.In the second step,retracking incorporates both the leading and trailing edges using the previously acquired a priori information.We tested the algorithm using data from the HY-2B radar altimeter.Results indicate that the proposed two-step retracking algorithm outperforms the Maximum Likelihood Estimation(MLE4)algorithm currently used in the operational processing of the HY-2B radar altimeter,as well as the Adaptive Leading Edge Subwaveform(ALES)algorithm,in terms of significant wave height(SWH)and sea level anomalies(SLA).Specifically,the standard deviation of the difference in SWH is reduced by 14%,and the standard deviation of the difference in SLA is reduced by approximately 18%.The two-step retracking algorithm effectively leverages trailing edge information,reduces the influence of peak noise on the leading edge,and improves both the utilization and accuracy of the waveform retracking.
基金financial supports from the National Natural Science Foundation of China-Youth Project(51801076)the Provincial Colleges and Universities Natural Science Research Project of Jiangsu Province(18KJB430009)+1 种基金the Postdoctoral Research Support Project of Jiangsu Province(1601055C)the Senior Talents Research Startup of Jiangsu University(14JDG126)。
摘要To solve the problems of deformation,micro-cracks,and residual tensile stress in laser cladding coatings,the technique of laser cladding with Fe-based memory alloy can be considered.However,the process of in-situ synthesis of Fe-based memory alloy coatings is extremely complex.At present,there is no clear guidance scheme for its preparation process,which limits its promotion and application to some extent.Therefore,in this study,response surface methodology(RSM)was used to model the response surface between the target values and the cladding process parameters.The NSGA-2 algorithm was employed to optimize the process parameters.The results indicate that the composite optimization method consisting of RSM and the NSGA-2 algorithm can establish a more accurate model,with an error of less than 4.5%between the predicted and actual values.Based on this established model,the optimal scheme for process parameters corresponding to different target results can be rapidly obtained.The prepared coating exhibits a uniform structure,with no defects such as pores,cracks,and deformation.The surface roughness and microhardness of the coating are enhanced,the shaping quality of the coating is effectively improved,and the electrochemical corrosion performance of the coating in 3.5%NaCl solution is obviously better than that of the substrate,providing an important guide for engineering applications.
摘要针对燃煤电厂参与调峰负荷波动较大,出口SO2浓度控制效果不佳的问题,建立了一种基于捕鱼优化算法(catch fish optimization algorithm,CFOA)优化融合神经网络的出口SO2浓度预测模型。首先使用互信息算法筛选由机理分析得到的特征变量,并通过逐次变分模态分解对筛选后的辅助变量进行分解重构,保留相关性较大的重构分量作为输入变量。随后采用双向时间卷积网络、双向门控循环单元与多头自注意力机制构建融合神经网络模型,通过CFOA对模型超参数寻优以进一步提高精度。最后使用某660 MW燃煤电厂历史运行数据进行对比实验,实验结果表明,该模型在出口SO2浓度剧烈波动的工况下仍能实现较好的预测效果。同多种模型对比,该模型具有更小的误差和更高的预测精度,体现出其在复杂变化环境中的鲁棒性和可靠性。