复杂车辆图像中的车牌定位与字符分割方法

车牌定位和字符分割是车牌照自动识别系统中的关键步骤。提出了一个综合多种特征的车牌定位算法和一个基于模板匹配的字符分割算法,有效地解决了背景复杂的彩色图像中车牌定位和字符分割的问题。综合这些算法并结合字符识别核心,实现了一个完整的车牌照自动识别系统。该系统对不同背景、光照条件下采集到的车辆图像进行了大量实验,实验结果证明算法准确率高、鲁棒性好。     

Substratum attachment location and biofilm formation by Bacillus cereus strains isolated from different sources: Effect on total biomass production and sporulation in different growth conditions

Bacillus cereus, an endospore forming human pathogen associated with foodborne diseases, can form biofilms and attach to surfaces of processing equipment in the food industry. It is a consistent source of contamination and/or cross contamination of processed food products. The objective of this study was to understand substratum attachment location and biofilm formation behavior of B. cereus strains under different growth conditions. A total of 60 strains isolated from food, human, or farm and a number of reference strains were used in this study. Substratum attachment locations of these strains in 96-well microtiter plates were highly diversified among these strains. Strains isolated from food showed higher preference to attach at the air-liquid interface during early stage of biofilm formation. To the best of our knowledge, this is the first report showing the level of substratum attachment location and biofilm formation of B. cereus strains isolated from different sources. Substratum properties did not affect biofilm formation location when a number of selected strains were grown on stainless steel coupon, indicating that biofilm formation location might be independent of the type of substratum. Substratum attachment location and biofilm formation related phenotypes such as total biomass production, number of sessile cells, and sporulation were closely correlated. Substratum attachment location and sporulation behavior were strongly affected during biofilm formation under nutrient stress condition. The number of spores was significantly increased in biofilms grown under nutrient stress condition even though total biomass formation was lower. Our results on substratum attachment location and related biofilm formation behavior are substantially important for food industries where different surfaces are prone to B. cereus attachment, particularly for setting up and implementing clean in place (CIP) protocols.     

基于信息综合判断的智能变电站网络通信故障定位技术研究

介绍了基于信息综合判断的智能变电站网络通信故障定位技术。它是以IEC61850标准中的变电站自动化拓扑结构、功能层和逻辑接口、ACSI模型的基本模型规范和信息交互模型规范、公用数据属性类型中的品质要求为基础,结合VLAN、Subnet等网络技术的应用情况,对智能变电站站控层、过程层网络状态、链路结构、网络接口位置及接口信息等进行系统组态、流程化管理,以实现对站控层、过程层设备网络信息的实时、准确管理,达到对智能变电站通信故障准确定位快速查找的目的。介绍的故障定位技术,在110 kV智能变电站进行了测试应用,测试结果表明达到了故障定位快速查找和有效评估网络故障可能带来的安全风险的目的。     

EDGE LOCALIZATION BY MULTIPLE OF GAUSSIANs THEORY

Zero-crossing of a derivative of Gaussian filter is a well-known edge location criterion. Examples are the Laplacian, the second derivative in the gradient direction (SDGD) and the sum of the Laplacian and SDGD (PLUS). Derivative operators can easily be implemented by convoluting the primitive image with a derivative of a Gaussian. Gaussian filter displaces the equipotential of half height inwards for convex edge and outwards for concave edges. A Difference-of-Gaussian (DoG) filter is similar to the Laplacian-of-Gaussian but with opposite sign and causes a convex edge shift inwards. This paper introduces the Multiple-of-Gaussian niters to reduce curvature-based location error. Using a linear combination of N Gaussians(N>2) with proper weights, the edge shifts can be reduced to 1/(2N-3) of the ones produced by a similar Laplacian-of-Gaussian filter.     

油田含CO2伴生气变压吸附材料实验研究

变压吸附工艺是国内小规模CO_2驱产出气处置的常用手段,而吸附剂是吸附分离的基础。选用6种沸石类吸附剂,测试了CO_2/CH4单组分吸附平衡等温线,分析了温度对吸附量的影响。优选了两种材料进行了CO_2-CH4混合气体竞争吸附实验,重点分析了温度、压力、气体组分、CO_2含量等因素对吸附剂吸附分离效果的影响,评价了油、水含量对吸附剂的损伤。实验结果表明:吸附剂的分离系数与混合气体中CO_2含量无明显相关性,但吸附体系压力的变化对混合气体分离效果影响较大。在初始气体CO_2摩尔分数50%、温度20℃、压力2.5MPa条件下,优选的吸附剂13X对混合气体的分离系数为25。油水的存在会对吸附剂的分离性能产生较大影响,在温度20℃、压力2.5 MPa、13X吸附剂含水摩尔分数10%及含油摩尔分数10%条件下,吸附总量分别下降32%和78%,分离系数分别下降48%和78%。     

误差敏感竞争性学习算法

本文基于等误差准则提出了一种适用于矢量量化技术的新型码书设计算法。实验表明此算法优于现存算法。为解决初始码书赋值问题，本文提出了自生成自组织神经网络方法。实验表明此算法加速了算法的收敛速度，提高了算法的性能     

能源互联网背景下基于势博弈的多供电主体竞争策略研究

能源互联网背景下,能源供需市场呈现更为显著的智能化和开放性,同时供电主体也逐步呈现多元化、竞争化态势.针对开放性市场环境下多个供电主体竞争并存的格局,分析了单个供电主体与其他供电主体的竞争互动问题,基于势博弈理论构建了多供电主体间的竞争博弈模型.同时,设计了动态互动电价,刺激用户与供电主体间的供需互动,形成以用户需求为...     

有杆抽油系统故障诊断的人工神经网络方法

将人工神经网络用于有杆抽油系统故障的自动识别.对江苏油田的实测示功图数据进行了预处理,利用Matlab6.5进行编程,应用相同的数据对BP神经网络模型和自组织竞争神经网络模型的识别效率进行了对比.结果表明,由自组织竞争神经网络建立的模型对测试数据的正确识别率更高,识别效果稳定.因此,将自组织竞争神经网络应用于示功图的自动识别问题对实现有杆抽油系统故障诊断的自动化以及实现真正意义上的数字油田提供了一种有效途径.     

铈离子引进方法对其在Y型分子筛中定位的影响

对离子交换法和一步沉积法引进铈离子的系列Y分子筛进行了铈离子定位研究。建立了铈离子在Y分子筛中定位的简便有效的基础表征体系：利用Ce3d XPS谱获得Ce4+/Ce3+摩尔比,通过Ce3d谱及XPS定量结果分析铈离子在分子筛笼内外的存在情况;利用XRD和Py IR确定铈离子在笼内的定位。研究了不同方法引进不同含量铈离子的定位规律。结果表明,离子交换法较沉积法利于铈离子迁移进分子筛笼内,从而进一步迁移进β笼;Ce2O3质量分数为4%、8%、12%和14%时,离子交换法引进的铈离子基本能迁移进β笼。一步沉积法引进的Ce2O3质量分数为4%时,易进入β笼;更多的铈离子则易留在分子筛笼外,可用于防止重金属中毒。     

一种肤色定位的人脸检测算法

进行人脸识别前,首先要精确定位出一幅图像中人脸的位置,为了快速定位人脸位置,本文提出一种肤色定位的人脸检测算法。首先将实际获取的彩色图像转换为YCbCr和HSI空间图像,并将Cb和Cr图像中的数值进行四舍五入处理,结合Cb、Cr、H和S的阈值去除大部分背景,再统计当前图像中的彩色部分的Cb和Cr值,分别取最多2个数值来共同确定肤色位置,最后由当前位置的亮度信息图像排除手等纯肤色部分,准确定位人脸。本文算法能提高正检率并降低误检率,有利于人脸检测。