基于模糊神经网络的地图匹配算法

提出了基于模糊神经网络的新的地图匹配算法.该算法综合了数字道路信息和GPS/DR定位信息,提取两个重要参数作为输入变量,即定位点到候选路段的投影距离及定位航向与候选路段方位角差.设计出了四层模糊神经网络及改进的收敛学习规则.实验结果表明所提出的算法能很好地匹配车辆行驶路段位置.      

基于复数卷积神经网络烧结台车定位跟踪算法

烧结台车定位跟踪是非稳态烧结高质量过程控制的基础,同时也是台车状态监控的关键。针对现有烧结生产过程中的物料跟踪和过程控制的问题,提出了一种全新的台车定位跟踪算法。该算法针对烧结台车定位场景设计了复数卷积神经网络结构,运用复数网络的俯角解决了相位从360°到0°跳变时定位系统误差大的问题;同时采用GradCAM技术解析网络的工作原理,以确保模型的泛化能力;通过在台车上加装定位二维码来二次确认定位信息的准确性。经过长达一年半的论证,证明系统具有高度的可靠性和稳定性,并能够有效地应用于烧结生产过程中的物料跟踪和过程控制。     

多媒体传感器网络中被动目标定位算法

针对被动目标定位问题,提出了一种多视觉感知节点协同定位算法.与传统的视觉目标定位方法不同,该算法无需进行训练或特征识别,而是各个节点基于机器视觉理论,将目标投影为定位平面上的一条直线,并使用Hough变换将该直线映射为Hough空间中的一个离散点,最后对离散点进行正弦曲线拟合,拟合参数即为目标的最终位置估计.根据视觉感知节点所处高度及分辨率引入置信度参数,进一步提高了多视觉感知节点协同定位算法精度.最后分析了算法的时间复杂度,并通过实验验证了该算法可以获得5 cm以上的定位精度.     

联合改进作用距离和Dijkstra算法的复杂结构声发射定位方法及应用

声发射定位对含空区等复杂结构的连续动态安全监测具有重要意义。针对直线路径定位方法不适用于复杂结构,而常规Dijkstra搜索算法常出现局部最优路径等问题,提出了一种联合改进作用距离和Dijkstra算法的声发射定位方法,实现复杂结构下声发射的高精度定位。理论测试显示：基于改进作用距离的Dijkstra算法得到的P波传播路径长度小于等于常规作用距离的Dijkstra算法,即P波走时更为准确。理论测试和断铅试验定位测试表明：本文提出的定位方法在复杂结构声发射定位时的整体误差在0.50 cm范围之内,断铅事件平均定位误差由常规作用距离Dijkstra算法的0.95 cm下降至本文的0.54 cm。改进的声发射定位方法在复杂结构声发射定位方面具有很好的应用前景。     

基于ALBERT与双向GRU的中医脏腑定位模型

脏腑定位，即明确病变所在的脏腑，是中医脏腑辨证的重要阶段。本文旨在通过神经网络模型搭建中医脏腑定位模型，输入症状文本信息，输出对应的病变脏腑标签，为实现中医辅助诊疗的脏腑辨证提供支持。将中医的脏腑定位问题建模为自然语言处理中的多标签文本分类问题，基于中医的医案数据，提出一种基于预训练模型ALBERT和双向门控循环单元（Bi-GRU）的脏腑定位模型。对比实验和消融实验的结果表明，本文提出的方法在中医脏腑定位的问题上相比于多层感知机模型、决策树模型具有更高的准确性，与Word2Vec文本表示方法相比，本文使用的ALBERT预训练模型的文本表示方法有效提升了模型的准确率。在模型参数上，ALBERT预训练模型相比BERT模型降低了模型参数量，有效减小了模型大小。最终，本文提出的脏腑定位模型在测试集上F1值达到了0.8013。      

基于半监督流形学习的无线传感器网络定位方法及其应用

提出了一种基于流形半监督学习的移动节点定位算法.该算法利用基于流形学习的半监督方法,通过一定量的有标签样本和无标签样本,获取隐含在节点接收信号强度信息中的流形结构,直接建立节点物理位置与接收信号强度之间的映射关系.算法不需要使用现有的理论或经验信号传播模型,避免了模型不准确带来的定位误差,而且允许网络中存在大量无标签样本,降低了数据采集难度,提高了算法实用性.冶金工业现场的实际应用结果表明,相对RADAR算法,本文算法具有较高的定位精度.     

基于编码器与NFC修正融合的带式输送机轨道式巡检机器人定位方法

轨道式巡检机器人的高精度定位技术是带式输送机智能化巡检的重要研究方向之一,而矿用带式输送机距离超长,工作环境复杂,严重影响巡检机器人的定位精度。针对目前的轨道式巡检机器人定位技术在矿用带式输送机巡检领域存在的问题,提出了基于编码器和NFC双传感器修正融合的高精度定位方法。分析带式输送机轨道式巡检机器人轨道与环境特性对编码器系数的影响,提出轨道分段原则。利用机器人搭载的编码器数据反馈特点,构建编码器递推定位方法。通过机器人运行的历史数据,对编码器系数进行分段分方向修正,并提出基于递推最小二乘的编码器系数修正方法,以提高编码器对轨道环境的适应性。在此基础上,根据机器人所在轨道分段的位置不同,在段端基于卡尔曼滤波算法实现编码器和NFC数据融合,在段内利用分段分方向修正系数与编码器信息进行递推定位,实现轨道式巡检机器人连续高精度的定位。针对所提方法搭建了实验平台并进行了实物测试,实验结果表明,相较于编码器定位、RFID定位和两者融合定位三种传统定位方式,基于编码器和NFC的修正融合定位算法能够有效提高轨道式巡检机器人定位对轨道环境的适应性,同时提高轨道式巡检机器人的定位精度。      

基于手机移动传感网络的矿震群智定位方法

为提高矿震监测系统定位精度,减少监测盲区,降低监测成本,基于分布式的思想,提出一种基于手机移动传感网络的矿震定位方法。首先以矿区附近工人及家属等使用的智能手机建立手机移动传感网络,其次对模拟震源点网格化,构建基于标准差的目标函数,提出改进的萤火虫寻优策略,并使用拐点回溯法以及手机移动传感网络排除离散点策略(EDPS)降低定位误差,最后通过矿震模拟实验进行验证。实验结果表明:在手机移动传感网络无到时误差理想情况下,所有模拟震源点都能够准确收敛至震源位置,定位误差小于1 m。但手机相较于检波器到时误差较高,且定位误差与到时误差具有相关性,当手机到时误差为?1.0～1.0 s时,传统算法定位误差为216 m,无法实现高精度定位。通过研究目标函数值与定位误差间的关系,提出并使用拐点回溯法以及EDPS两种优化方法,算法绝对定位误差降低至73 m,当到时误差为?0.2～0.2 s时,绝对定位误差降低至17 m,定位精度提高76.1%。基于手机移动传感网络的矿震群智定位方法,为矿震监测提供了一种新方法,未来可考虑与井下微震系统联合,在节省监测成本、提高定位精度方面具有重要意义。      

视线追踪中一种新的由粗及精的瞳孔定位方法

基于图像处理的方法,采用了由粗及精的瞳孔定位思想,提出了一种高精度的瞳孔定位算法。该算法首先利用瞳孔区域的直方图,采用改进的最大类间方差法自适应地分割瞳孔区域,实现粗略定位,然后利用瞳孔灰度的梯度特性来精确定位瞳孔边缘点,最后在像素级瞳孔边缘点的基础上,采用亚像素定位方法,更精确地求得亚像素级瞳孔边缘点,并通过椭圆拟合的方法来精确确定瞳孔的中心位置。另外,针对瞳孔被遮挡的情况,本文提出了一种等距离补偿瞳孔的方法。多次实验结果证明了该算法对遮挡瞳孔的定位有较强的鲁棒性,可以准确地定位瞳孔的位置。      

基于UWB的地下定位算法和拓扑优化

地下定位面对环境恶劣、干扰、多径等影响,常规算法难以获得高精度的定位结果,同时井下环境多为狭长的巷道,不利于布置定位所需的锚节点,而井下锚节点的布置通常对定位结果有较大影响,因而使用普通的定位方法不足以满足智能采矿所需的高精度定位需求.本文对传统的三边定位算法进行分析,总结了传统三边定位结果产生误差的原因,并提出了改进的算法,通过仿真实验验证了改进算法的有效性.同时通过理论分析误差带,使用最大绝对定位误差用于仿真分析拓扑结构对定位结果精度的影响,提出了对拓扑结构的优化原则,能够根据环境特点以实现定位区域内平均最大绝对定位误差最小为原则得出最优拓扑结构.文中设置了仿真实验和实地实验对改进的算法和拓扑结构优化方法进行了验证,实验结果中,改进的算法能够在相同拓扑结构下减小15%~43%的误差,而在相同算法下优化的拓扑结构能够减小17%~65%,二者结合能够减小误差达74%.结果表明,在相同的定位条件下,改进的定位算法能够明显提高定位结果的精度,同时定位结果与拓扑结构之间也有着密切的联系,根据实际环境灵活布置拓扑结构能够使定位结果的精度进一步提高,将改进的算法与拓扑结构优化方法结合可以实现更高的定位精度.      

Reliability and validity of the measurement of position sense in women with osteoarthritis of the knee

OBJECTIVE: We examined the test-retest reliability and the construct validity of the measurement of knee position sense for describing the functional weightbearing performance of women with osteoarthritis (OA) of the knee. METHODS: For the purpose of this study, position sense was defined as the error occurring when subjects attempted to reproduce a criterion angle in standing with visual cues eliminated. Five such tests were recorded photographically on 3 different occasions. On each occasion the 10 subjects also completed a self-paced walking test over a 13 m indoor walkway. RESULTS: The photographic measurements were reproducible (r = 0.90) and there was no change in positioning accuracy across sessions. There was a significant (p < 0.05) inverse correlation of 0.70 between the standard deviation of the mean individual measurements of position sense (precision of the test) and those of walking speed. CONCLUSION: Our study demonstrates good measurement reliability and a comparable mean angular error with repeated tests. It also suggests the amplitude of the variability of this error is a strong determinant of an individual OA patient's functional performance in walking.     

Demand Forecasting for Irrigation Water Distribution Systems

One of the main problems in the management of large water supply and distribution systems is the forecasting of daily demand in order to schedule pumping effort and minimize costs. This paper examines methodologies for consumer demand modeling and prediction in a real-time environment for an on-demand irrigation water distribution system. Approaches based on linear multiple regression, univariate time series models (exponential smoothing and ARIMA models), and computational neural networks (CNNs) are developed to predict the total daily volume demand. A set of templates is then applied to the daily demand to produce the diurnal demand profile. The models are established using actual data from an irrigation water distribution system in southern Spain. The input variables used in various CNN and multiple regression models are (1) water demands from previous days; (2) climatic data from previous days (maximum temperature, minimum temperature, average temperature, precipitation, relative humidity, wind speed, and sunshine duration); (3) crop data (surfaces and crop coefficients); and (4) water demands and climatic and crop data. In CNN models, the training method used is a standard back-propagation variation known as extended-delta-bar-delta. Different neural architectures are compared whose learning is carried out by controlling several threshold determination coefficients. The nonlinear CNN model approach is shown to provide a better prediction of daily water demand than linear multiple regression and univariate time series analysis. The best results were obtained when water demand and maximum temperature variables from the two previous days were used as input data.     

Thermal comfort in offices:comfort values and optimization of indoor control variables

In order to search for reasonable air-conditioned indoor control variables and save energy consumption and meet tO need of personal thermal comfort,a method which is based on numerical simulation is employed to optimize indoor control variables.Computational fluid dynamics(CFD)is used to describe thermal state of office.An optimal method is proposed in this paper,dual neural network model is firstly used to acquire reliable information,data from CFD model are pre-processed,and the remaining data are used to train artificial neural networks(ANN),then CFD model is replaced by ANN model to reduce computational cost when is optimized,indoor control variables are optimized by genetic algorithm.Simulation results show that indoor thermal comfort is improved obviously,and the energy cost is decreased accordingly.     

Robust Optimal Design of Activated Sludge Bioreactors

This paper proposes an algorithm for a robust optimal design of the biological reactor and secondary settling facilities in suspended growth nitrogen and phosphorus removal systems. Robust optimization includes uncertainty in the decision-making procedure and seeks a solution that remains “close” to optimal for all potential operation scenarios. It thus differs fundamentally from the deterministic and stochastic approaches, where uncertainty is ignored or a solution based on either the most likely scenario or the average performance over all potential scenarios is produced. The robust optimization of a suspended growth system is a multiobjective optimization problem concerned with minimization of the global costs and variability of the system’s performance around the optimal. The proposed robust optimization approach uses the ASM3 model, making use of its performance prediction capabilities to produce a powerful tool for designing activated sludge systems. The algorithm was applied to the design of the biological reactor and secondary settling facilities for the Vila Real municipal wastewater treatment plant (Portugal).     

基于YOLOv3的无人机识别与定位追踪

近年来，无人机入侵的事件经常发生，无人机跌落碰撞的事件也屡见不鲜，在人群密集的地方容易引发安全事故，所以无人机监测是目前安防领域的研究热点。虽然目前有很多种无人机监测方案，但大多成本高昂，实施困难。在5G背景下，针对此问题提出了一种利用城市已有的监控网络去获取数据的方法，基于深度学习的算法进行无人机目标检测，进而识别无人机，并追踪定位无人机。该方法采用改进的YOLOv3模型检测视频帧中是否存在无人机，YOLOv3算法是YOLO（You only look once，一次到位）系列的第三代版本，属于one-stage目标检测算法这一类，在速度上相对于two-stage类型的算法有着明显的优势。YOLOv3输出视频帧中存在的无人机的位置信息。根据位置信息用PID（Proportion integration differentiation，比例积分微分）算法调节摄像头的中心朝向追踪无人机，再由多个摄像头的参数解算出无人机的实际坐标，从而实现定位。本文通过拍摄无人机飞行的照片、从互联网上搜索下载等方式构建了数据集，并且使用labelImg工具对图片中的无人机进行了标注，数据集按照无人机的旋翼数量进行了分类。实验中采用按旋翼数量分类后的数据集对检测模型进行训练，训练后的模型在测试集上能达到83.24%的准确率和88.15%的召回率，在配备NVIDIA GTX 1060的计算机上能达到每秒20帧的速度，可实现实时追踪。      

基于深度学习的高效火车号识别

基于高性能的YOLOv3目标检测算法,提出一种分阶段高效火车号识别算法。整个识别过程分为两个阶段:第一阶段在低分辨率全局图像中检测出火车号区域位置;第二阶段在局部高分辨率图像中检测出组成火车号的字符,根据字符的空间位置关系搜索得到12位火车号,并利用每个字符的识别置信度及火车号编码规则进行校验得到最终火车号。另外,本文提出一种结合批一化因子和滤波器相关度的剪枝算法,通过对两个阶段检测模型的剪枝,在保证识别准确率不降（实验中略有提升）的条件下降低了存储空间占用率和计算复杂度。在现场采集的1072幅火车号图像上的实验结果表明,本文提出的火车号识别算法达到了96.92%的整车号识别正确率,平均识别时间仅为191 ms。      

Modeling Particle Dispersion under Human Activity Disturbance in a Multizone Indoor Environment

Human activity is an important factor influencing particle resuspension in the indoor environment. This work studies the applicability of a multizone airflow and contaminant transport model (CONTAM 2.1) in the simulation of indoor dispersion of particles under human activity disturbance. An iterative method is suggested to complement CONTAM 2.1, by tracking the transient particle concentration on floor surface due to dynamic deposition and resuspension process. A three-zone building with a heating, ventilation, and air conditioning system is used as the simulation case to test the convergence and accuracy of this algorithm under different particle-release scenarios. The algorithm shows a very fast convergence speed in the simulation. Comparisons of calculation results between the multizone model and the analytical model show good agreement and verify the accuracy of the multizone model simulation. The airborne particle concentration profiles and human breathing dose are also analyzed for the three-zone building model.     

Wave-like reaction phenomena of the human body in head-down tilt

The biorhythmological analysis of the data obtained from 7 healthy male test subjects in the head-down position at -8 degrees revealed phenomena of sine-shaped reactions. This pattern was seen in the range (the difference between mean day-time and mean night-time values) of heart rate, amplitude (the difference between the maximum and minimum) of variations of body temperature and renal potassium excretion, and the evening-morning difference of body temperature variations.