基于本体和知识组件的高低机设计方法研究

为了提高高低机设计质量和效率,提出一种基于本体和知识组件的高低机设计方法。利用本体和知识组件开展高低机设计知识的表示与建模,构建高低机设计知识库,在明确基于本体和知识组件的高低机设计方法的基础上,以Java为平台开发了高低机设计专用系统,并以某高低机的设计实例进行验证。验证结果表明：该方法及开发的设计系统具有可行性和有效性,能够按照本体中既定设计流程,依次调用对应知识组件,实时提示相关设计知识,辅助设计人员完成繁琐与重复的计算工作。     

巡飞弹航迹导引方法研究

针对巡飞弹等高、常速巡飞等特点,将整条航迹分为直线段和转弯段,利用假设条件,提出一种导引方法:在直线段进行位置和偏差角控制,转弯段采用转弯控制,进而完成整个航迹导引并确保巡飞弹在飞行过程中偏离航线的误差较小。通过仿真可以看出,此方法使得巡飞弹在整个飞行过程中偏离航线的误差极小。     

基于数字图像处理的弹目偏差量测量的研究

弹目偏差量的测量是高炮武器系统校射的关键,为了提高其测量的实时性和测量精度,文中采用DSP进行高速数字图像处理,研究图像处理技术和弹目偏差量实时获取的方法.图像处理的方法进行弹目偏差量的测量,从根本上避免了人员目视测量,具有较高的测量精度,为高炮武器系统校射精度的提高奠定基础.     

采用蒸馏训练的时空图卷积动作识别融合模型

目的 基于深度学习的动作识别方法识别准确率显著提升,但仍然存在很多挑战和困难。现行方法在一些训练数据大、分类类别多的数据集以及实际应用中鲁棒性较差,而且许多方法使用的模型参数量较大、计算复杂,提高模型准确度和鲁棒性的同时对模型进行轻量化仍然是一个重要的研究方向。为此,提出了一种基于知识蒸馏的轻量化时空图卷积动作识别融合模型。方法 改进最新的时空卷积网络,利用分组卷积等设计参数量较少的时空卷积子模型;为了训练该模型,选取两个现有的基于全卷积的模型作为教师模型在数据集上训练,在得到训练好的教师模型后,再利用知识蒸馏的方法结合数据增强技术训练参数量较少的时空卷积子模型;利用线性融合的方法将知识蒸馏训练得到的子模型融合得到最终的融合模型。结果 在广泛使用的NTU RGB + D数据集上与前沿的多种方法进行了比较,在CS(cross-subject)和CV(cross-view)两种评估标准下,本文模型的准确率分别为90.9%和96.5%,与教师模型2s-AGCN(two-stream adaptive graph convolutional networks for skeleton-based action)相比,分别提高了2.4%和1.4%;与教师模型DGNN(directed graph neural network)相比,分别提高了1.0%和0.4%;与MS-AAGCN(multi-stream attention-enhanced adaptive graph convolutional neural network)模型相比,分别提高了0.9%和0.3%。结论 本文提出的融合模型,综合了知识蒸馏、数据增强技术和模型融合的优点,使动作识别更加准确和鲁棒。     

超声测流在渐缩流道中的应用及其准确度分析

多声道超声测流装置用于圆形管道或矩形方涵时具有成熟的数学模型,用于泵站入口的渐缩流道时,由于流道截面积的变化和内部流速分布的复杂性,大大提升了流量测量的难度。本文利用CFD模拟了前柳林泵站渐缩流道的内部流场,分析了超声测流装置的声道流速分布特征,比较了渐缩流道中两种不同流量计算模型的测量准确度。结果表明,渐缩流道中应用矩形流道算法导致的流量计算误差大约在3% ~ 4%之间,矩形优化算法OWIRS的准确度提升并不适用于渐缩流道,通过修正G-L方法计算截面的位置可以提升测量准确度。实际测量中声道缺失会导致明显的流量系统偏差,利用曲线拟合的方法可以在一定程度上使偏差得到补偿。     

一种电站锅炉热力计算的分析与修正方法

通过对烟气侧热偏差的产生原因进行深入分析,发现辐射换热是一个易被研究人员忽视的因素。结合某660 MW超超临界燃煤机组,在热力校核计算基础上,提出将水平烟道分为左、右两部分独立分析。分别定义偏差系数M、N,建立计算模型,探索辐射换热偏差对位于炉膛出口的末级过热器汽温偏差的影响。结果表明,在相同M、N值下,低负荷汽温偏差略大于高负荷,且汽温偏差值随M、N的增加呈近线性递增。     

一种基于ARMA-BP组合模型的电压偏差预测方法

国内许多变电站建立了电能质量的预测与预警机制,以应对日益严重的电能质量问题,其中电压偏差最为严重。针对预测模块中电压偏差预测算法的缺失,结合配电网的运行状态,提出了一种基于ARMA-BP组合模型的电压偏差预测方法。针对单一时间序列方法的不足,将时间序列和人工神经网络的算法结合起来。通过分析上海某变电站的电压偏差数据特征,首先采用时间序列的方法建立ARMA模型。然后采用BP人工神经网络的方法对ARMA模型预测值与原始数据之间的残差值进行拟合预测,最终得到2种模型预测所得累加值的结果。研究结果表明了所提方法的有效性。     

建筑企业内部知识共享影响因素的实证研究

通过建立结构方程模型,对工程项目环境下影响建筑企业内部知识共享的因素进行实证研究。研究表明,项目经理支持及同事关系对促进组织内部正式和非正式知识共享均具有显著作用;共享机制能够显著促进正式知识共享,但对非正式知识共享的直接影响作用并不明显,需通过同事关系间接影响非正式知识共享;组织激励虽然能够显著促进正式知识共享,但对非正式知识共享却有一定的抑制作用。     

Place Making for Knowledge Generation and Innovation: Planning and Branding Brisbane's Knowledge Community Precincts

Knowledge generation and innovation have been a priority for global city administrators particularly during the last couple of decades. This is mainly due to the growing consensus in identifying knowledge-based urban development as a panacea for burgeoning economic problems. Place making has become a critical element for success in knowledge-based urban development as planning and branding places is claimed to be an effective marketing tool for attracting investment and talent. This paper aims to investigate the role of planning and branding in place making by assessing the effectiveness of planning and branding strategies in the development of knowledge and innovation milieus. The methodology of the study comprises reviewing the literature thoroughly, developing an analysis framework, and using this framework to analyze Brisbane's knowledge community precincts—namely Boggo Road Knowledge Precinct, Kelvin Grove Urban Knowledge Village, and Sippy Downs Knowledge Town. The analysis findings generate invaluable insights into Brisbane's journey in place making for knowledge and innovation milieus and communities. The results suggest as much as good planning, branding strategies, and good practice, the requirements of external and internal conditions also need to be met for successful place making in knowledge community precincts.     

Application of ontology in vulnerability analysis of metro operation systems

Vulnerability analysis plays a key role in the metro operation system, since the metro operation security has become more important than ever. The vulnerability management is a systems engineering, it covers different departments of the metro operation company. A successful vulnerability management needs the close cooperation and coordination between them. However, it has blocked a successful vulnerability management due to lack of the common knowledge basis. This paper applies ontology into the vulnerability analysis of the metro operation system and illustrated the major design ideas of the ontology model. Examples are given as well to demonstrate the implementation of knowledge reasoning based on the ontology. The ontology was also evaluated through expert evaluation for further improvement. By using this technology, a common knowledge base of the vulnerability management in the metro operation system can be built. It provides information sharing, communication and collaboration for all the participants in the system for decision-making. Moreover, further development of the intelligent application can be built based on it, efficiently increasing the management level of the vulnerability.