Aiming at the requirement for high-precision tilt monitoring in the field of structural health monitoring(SHM),this paper proposes a sensitivity-enhanced tilt sensor based on a femtosecond fiber Bragg grating(FBG).Fir...Aiming at the requirement for high-precision tilt monitoring in the field of structural health monitoring(SHM),this paper proposes a sensitivity-enhanced tilt sensor based on a femtosecond fiber Bragg grating(FBG).Firstly,structural design of the tilt sensor was conducted based on static mechanics principles.By positioning the FBG away from the beam’s neutral axis,linear strain enhancement in the FBG was achieved,thereby improving sensor sensitivity.The relationship between FBG strain,applied force,and the offset distance from the neutral axis was established,determining the optimal distance corresponding to maximum strain.Based on this optimization scheme,a prototype of the tilt sensor was designed,fabricated,and experimentally tested.Experimental results show that the FBG offset distance yielding maximum sensitivity is 4.4 mm.Within a tilt angle range of−30°to 30°,the sensor achieved a sensitivity of 129.95 pm/°and a linearity of 0.9997.Compared to conventional FBG-based tilt sensors,both sensitivity and linearity were significantly improved.Furthermore,the sensor demonstrated excellent repeatability(error<0.94%),creep resistance(error<0.30%),and temperature stability(error<0.90%).These results demonstrate the sensor’s excellent potential for SHM applications.The sensor has been successfully deployed in an underground pipeline project,conducting long-term monitoring of tilt and deformation in the steel support structures,further proving its value for engineering safety monitoring.展开更多
针对目前原始自适应蒙特卡洛定位(Adaptive Monte Carlo Localization,AMCL)在相似环境下绑架检测容易出错且重定位极易失败等问题,提出基于墙角族语义尺寸链的改进AMCL算法.融合机器人多传感器信息和Gmapping算法构建二维栅格地图,基于...针对目前原始自适应蒙特卡洛定位(Adaptive Monte Carlo Localization,AMCL)在相似环境下绑架检测容易出错且重定位极易失败等问题,提出基于墙角族语义尺寸链的改进AMCL算法.融合机器人多传感器信息和Gmapping算法构建二维栅格地图,基于Yolov5获取室内环境的目标检测框和类别信息,结合GrabCut算法和贝叶斯方法构建增量式语义映射地图;通过墙角的凸、凹和墙角相对于机器人的方位角对墙角进行分类,充分发掘语义映射地图中各墙角之间、墙角与室内物体之间的类别和位置关系,构建墙角族语义尺寸链和相应检索表;在定位过程中,基于墙角族语义尺寸链进行全局预定位,提出绑架检测机制进行绑架检测,在检测到绑架事件发生后,基于改进AMCL算法实现定位自恢复.最后,通过真实环境下的绑架实验验证了本文方法的有效性,实验表明,所提方法的全局定位准确率、全局定位速率、绑架检测准确率和绑架后定位准确率在相似环境下分别提升了42%、214%、88%和72%;在非相似环境下分别提升了44%、152%、12%和92%;在长走廊环境下分别提升了36%、426%、26%和68%.展开更多
按失效形式对大规模废旧零/部件进行预分类,是提高废旧零/部件批量再制造效率与效益的重要保障。针对大批量废旧零/部件表面失效人工识别效率低、漏检率和错检率高,导致难以满足自动化在线检测与分类需求的问题,提出一种基于机器视觉的...按失效形式对大规模废旧零/部件进行预分类,是提高废旧零/部件批量再制造效率与效益的重要保障。针对大批量废旧零/部件表面失效人工识别效率低、漏检率和错检率高,导致难以满足自动化在线检测与分类需求的问题,提出一种基于机器视觉的废旧零/部件批量在线表面失效形式识别与分类方法。在分析再制造检测服务概念与废旧零/部件失效形式的基础上,针对图像视觉下废旧零/部件“近形-异类”表面失效形式误判率高的问题,利用ROI高斯学习策略对废旧零/部件表面失效区域精准定位,提取候选分类特征,利用遗传算法(Genetic Algorithm,GA)筛选出关键分类特征,采用支持向量机(Library for Support Vector Machines,LIBSVM)建立失效形式分类模型,并通过K折交叉验证方法(K-fold Cross-Validation,K-CV)对其惩罚因子和核参数进行优化。以某退役齿轮零件为例对该方法的有效性与可行性进行验证,结果显示:该方法对再制造回收零/部件失效形式的分类精度达到96.7%,比同类算法精度提高了2.3%,比熟练人工检测精度提高了2.5%,表明该方法不仅具有一定的理论优越性,而且具有广阔的应用前景。展开更多
文摘Aiming at the requirement for high-precision tilt monitoring in the field of structural health monitoring(SHM),this paper proposes a sensitivity-enhanced tilt sensor based on a femtosecond fiber Bragg grating(FBG).Firstly,structural design of the tilt sensor was conducted based on static mechanics principles.By positioning the FBG away from the beam’s neutral axis,linear strain enhancement in the FBG was achieved,thereby improving sensor sensitivity.The relationship between FBG strain,applied force,and the offset distance from the neutral axis was established,determining the optimal distance corresponding to maximum strain.Based on this optimization scheme,a prototype of the tilt sensor was designed,fabricated,and experimentally tested.Experimental results show that the FBG offset distance yielding maximum sensitivity is 4.4 mm.Within a tilt angle range of−30°to 30°,the sensor achieved a sensitivity of 129.95 pm/°and a linearity of 0.9997.Compared to conventional FBG-based tilt sensors,both sensitivity and linearity were significantly improved.Furthermore,the sensor demonstrated excellent repeatability(error<0.94%),creep resistance(error<0.30%),and temperature stability(error<0.90%).These results demonstrate the sensor’s excellent potential for SHM applications.The sensor has been successfully deployed in an underground pipeline project,conducting long-term monitoring of tilt and deformation in the steel support structures,further proving its value for engineering safety monitoring.
文摘针对目前原始自适应蒙特卡洛定位(Adaptive Monte Carlo Localization,AMCL)在相似环境下绑架检测容易出错且重定位极易失败等问题,提出基于墙角族语义尺寸链的改进AMCL算法.融合机器人多传感器信息和Gmapping算法构建二维栅格地图,基于Yolov5获取室内环境的目标检测框和类别信息,结合GrabCut算法和贝叶斯方法构建增量式语义映射地图;通过墙角的凸、凹和墙角相对于机器人的方位角对墙角进行分类,充分发掘语义映射地图中各墙角之间、墙角与室内物体之间的类别和位置关系,构建墙角族语义尺寸链和相应检索表;在定位过程中,基于墙角族语义尺寸链进行全局预定位,提出绑架检测机制进行绑架检测,在检测到绑架事件发生后,基于改进AMCL算法实现定位自恢复.最后,通过真实环境下的绑架实验验证了本文方法的有效性,实验表明,所提方法的全局定位准确率、全局定位速率、绑架检测准确率和绑架后定位准确率在相似环境下分别提升了42%、214%、88%和72%;在非相似环境下分别提升了44%、152%、12%和92%;在长走廊环境下分别提升了36%、426%、26%和68%.
文摘按失效形式对大规模废旧零/部件进行预分类,是提高废旧零/部件批量再制造效率与效益的重要保障。针对大批量废旧零/部件表面失效人工识别效率低、漏检率和错检率高,导致难以满足自动化在线检测与分类需求的问题,提出一种基于机器视觉的废旧零/部件批量在线表面失效形式识别与分类方法。在分析再制造检测服务概念与废旧零/部件失效形式的基础上,针对图像视觉下废旧零/部件“近形-异类”表面失效形式误判率高的问题,利用ROI高斯学习策略对废旧零/部件表面失效区域精准定位,提取候选分类特征,利用遗传算法(Genetic Algorithm,GA)筛选出关键分类特征,采用支持向量机(Library for Support Vector Machines,LIBSVM)建立失效形式分类模型,并通过K折交叉验证方法(K-fold Cross-Validation,K-CV)对其惩罚因子和核参数进行优化。以某退役齿轮零件为例对该方法的有效性与可行性进行验证,结果显示:该方法对再制造回收零/部件失效形式的分类精度达到96.7%,比同类算法精度提高了2.3%,比熟练人工检测精度提高了2.5%,表明该方法不仅具有一定的理论优越性,而且具有广阔的应用前景。
