爆破拆除楼房时塌落振动的预测

从十余栋楼房塌落振动综合实测中可见,峰值振速总是由楼房倒塌触地的姿态所决定,即可能发生在后支撑爆破楼房下坐、切口闭合或翻倒触地时,且峰值也不尽相同。由此,在量纲分析中引入重心下落高度,分别建立了楼房下坐、楼房切口闭合冲击和建(构)筑物整体翻倒触地振动的峰值振速经验公式,阐述了相应的楼房、建(构)筑物不同塌落振动原理,并从案例实测振速对数图峰值最大包络线中,摄取公式待定参数Kt、β,由此分别提出对应的峰值振速算法的计算公式,并阐明参数的物理意义和取值。预测地点振速可先按结构选取,高大烟囱、现浇剪力墙(包括前跨现浇剪力墙的框剪结构及13层以上的单向倾倒现浇框剪结构的前方预测点)的塌落峰值振速,选取算法(3)计算峰值振速。框架和其他框剪楼房塌落振动的峰值,可按触地姿态选取算法(1)和算法(2)计算,并选取算法中的较大计算值为预测的峰值振速。由于补充了算法(1)和算法(2),综合算法正确地反映了形成峰值振速的楼房触地位置和撞地冲击时重心改变的高度,因此振动原理较明确,由此提高了预测塌落振动的针对性和准确性。并结合观测实例进行了公式验证,证明了公式的合理性。     

基于北斗导航通信技术的无人艇运动导航控制系统设计

为实现对无人艇行进器的运动控制,建立更加完善的导航应用策略,设计基于北斗导航通信技术的无人艇运动导航控制系统。以无人艇运动姿态研究作为切入点,在电机驱动器、电机测速模块的支持下,连接中央处理单元、电源模块、方位监测模块等硬件设备结构体,完成导航控制系统的硬件电路设计。在此基础上,确定与导航信息相关的短报文传输特征,通过协调接口数据传输协议应用格式的方法,定义通信协议加密对象的所属连接形式,完成对报文加密密钥参数的初步配置,实现基于北斗导航通信技术的无人艇运动导航控制系统的短报文加密处理。与基于惯性测量技术的导航控制系统相比,基于北斗导航通信技术控制系统的实际应用续航时间更长,在行进过程中始终能对无人艇运动方向进行精准化控制,可在完善行进器导航应用策略方面起到较强促进作用。     

小型断路器柔性装配中视觉识别系统的设计与应用

现代制造业对小型断路器（MCB）生产过程的效率和精度要求都在不断提高,传统的人工装配效率低且装配质量参差不齐,而传统基于振动盘上料的自动装配技术限制了制造的柔性化水平。针对上述问题以及未来的市场需求,提出了一种基于机器视觉的小型断路器柔性装配系统,该系统搭建专用的视觉识别模块,通过VGG-16架构的深度学习分类器和特征模板匹配方法,对小型断路器零件的种类、位置坐标、当前姿态进行识别,并将识别结果发送给工业机器人控制器,指导工业机器人对不同型号产品的不同零件类型通过机器人夹爪的灵活切换来完成不同的装配任务。实验表明,该系统对零件种类识别准确率为99.8%,坐标偏差在±0.3mm以内,旋转角度偏差在±0.8°以内,达到了MCB装配的精度要求,符合柔性化制造的需求。     

基于多目摄像机的鲁棒眼动跟踪技术研究

为改进单个远程传感器采集眼动数据时存在视场小、容易被遮挡物遮挡的缺点,研究了一种多目摄像机眼动跟踪技术,以更好地采集眼动数据。应用多个摄像机对人体头部姿态和眼球信息进行采集,通过分割视频帧提取瞳孔图像和计算被测用户头部姿态角度,将瞳孔图像放入卷积神经网络进行训练得到注视点坐标,并基于头部姿态信息计算每个摄像机的注视点权重,从而加权融合得到更精确的注视点信息。研究结果表明：在头部姿态角较大时,多目眼动追踪技术的精度比单目传感器的精度高30%~50%。该技术具有灵活性和通用性,在驾驶舱设计、用户用眼习惯评估和驾驶学员眼动绩效分析中具有重要的应用和推广价值。     

Assessment of gastroesophageal reflux disease,musculoskeletal symptoms and quality of life in dentists

Background and aimGastroesophageal reflux disease (GERD) has been associated with sick leave, decreased work productivity and poor quality of life. Another possible cause for the development of GERD is the relationship with maintaining the posture in flexion of the spine. The aim of this study was: (1) to compare gastro esophageal reflux with musculoskeletal disorders and quality of life of dentists, (2) to evaluate the association between gastro esophageal reflux such as time in the profession, time working while sitting, musculoskeletal disorders and quality of life of dentists.Methods83 dentists (23 males and 60 females; average age 39.3 years old) participated in the study, who were evaluated for their time in the profession, sitting time, musculoskeletal disorders, quality of life (SF-12v2) and the GERD symptoms using the Frequency Scale for Symptoms of GERD (FSSG) questionnaire.ResultsThe results showed that the majority were female (72.3%) and sendentary (74.7%). A difference was found between the level of GERD with musculoskeletal disorders (p < 0.05) and quality of life (p < 0.05). An association was also found between GERD and working time (p < 0.01), sitting working time (p < 0.05), headache (p < 0.01) and quality of life (p < 0, 01).ConclusionThis study showed that there was an association with working time, sitting time, musculoskeletal disorders, quality of life and GERD in dentists.     

基于柔性力敏传感器的便携式坐姿监测与提醒系统

读写坐姿长时间不规范将对青少年身体健康、学习效率产生不利影响.本文基于柔性力敏传感器,设计了一套新型的便携式读写坐姿实时监测及智能提醒系统.该系统根据生物力学、人体工程学规律,通过铺设于座椅上40 cm×40 cm的柔性力敏传感器,采集坐立时臀部压力分布数据。首先对原始数据进行预处理,再通过所提出的坐姿动作识别与分析算法对坐姿状态进行识别,实时分析坐姿情况,最后结合环境的温度、湿度、光强参数,给出预警方案.同时,在手机APP客户端实时显示并且存储监测期间坐姿及环境变化情况.该系统设计简单、交互性强,借助手机和压力垫的便携性,方便用户达到实时预警和分析的效果,帮助青少年纠正不良坐姿.     

基于足底压力的人体姿态检测和行为分析方法

人体的运动姿态检测和行为分析具有广泛应用价值.设计了一种基于足底压力的人体姿态检测和行为分析系统,系统由压力采集模块、数据处理模块、无线通信模块和行为分析软件组成.姿态检测系统获取足底压力数据,采用蓝牙方式进行数据通信,行为分析软件使用支持向量机多分类方法实现坐、站、走、跑和爬楼5个经典人体姿态的区分.实验证明该系统对人体姿态具有较好的识别精度和可靠性,可以用于人体姿态检测和行为分析.     

A survey of leaning and bracing behaviours in an automotive assembly plant

It is common for workers to use objects within the task environment to externally support themselves, when performing a task. We refer to this as ‘leaning’, when using the contralateral hand for support, or ‘bracing’, when using the trunk, hips or thighs for support. However, currently, ergonomists are unable to predict when these behaviours will occur when performing proactive ergonomic assessments, leading to incorrect estimates of how the jobs will be performed. The purpose of this survey was to quantify the types of external support used by workers during automotive assembly tasks. 250 automotive assembly jobs were observed, and the external support behaviours were documented using a 13-item survey adapted from Jones et al. (2008). We observed that 103 jobs (41.2%) were performed with the use of leaning and/or bracing. Of those, workers leaned, braced, or used both behaviours simultaneously 54.5%, 22.0%, and 23.5% of the time, respectively. Leaning was most prevalent at a height of 150 cm, while bracing was most prevalent at a height of 50 cm. This survey catalogues the leaning and bracing behaviours from a large array of assembly jobs. The trends we have identified can be used by ergonomists to predict whether or not external support is likely to be used by a worker during a certain task.     

一种精密三坐标POGO柱设计与精度研究

针对飞机数字化装配中大部件调姿与对合问题,提出一种由3个具有三平动自由度的分支机构构成并联机构对飞机大部件进行支撑和姿态调整的方法.根据飞机大部件调姿对合装配工艺要求,三坐标POGO柱应具有高的设计指标,对POGO柱结构进行详细设计,并提出用于提高其工作性能的关键措施.通过综合分析引起POGO空间定位误差的因素,建立空间定位误差的数学模型,进而得到确定导轨安装基准面形位公差的方法.实验结果表明,基于3个精密三坐标POGO柱的大部件调姿方法能够满足飞机部件调姿的高精度、高刚度要求.将精密三坐标POGO柱作为数字化柔性装配系统的基础部件,通过模块化配置还可以适应各类大型部件姿态调整的应用需求.     

基于RBF神经网络的人体动态姿态识别算法

利用三轴加速度传感器对人体姿态加速度信号进行采集和处理,通过径向等函数（RBF）神经网络算法对数据进行训练和分类,从而达到对人体基本动态姿态的识别。实验表明,该算法可以达到90％以上的识别率。