一座小型建筑空调动态负荷的ANN预测研究

人工神经网络(Artificial Neural Network,简称ANN)是一种崭新非线性建模和预测方法,具有良好的非线性品质和极高的拟合精度。在前人研究的基础上,针对一座小型建筑物建立了其空调动态负荷预测的BP模型。为克服常规BP训练算法的缺陷,提出了一种改进的遗传搜索算法,结果证明是一种高效实用的算法。     

雷达组网区域性防御区优化部署

对雷达网的费效比、数量、类型方面的优化部署进行了讨论。针对自卫压制式干扰或随队压制式干扰,给出一种采用不同频率的雷达利用不同位置进行优化部署的方法,并利用数据融合技术对这种部署方法进行讨论。     

无人机飞行训练资源配置效率的AHP-Arena综合评价方法

为降低传统层次分析法(analytic hierarchy process,AHP)赋权过程主观片面性影响,提出了一种综合AHP与Arena仿真建模技术的无人机训练资源配置效率评价方法。基于飞行训练流程分析,确定了训练资源配置效率影响因素和效率评价指标,建立了层次型综合评价体系,概括了AHP-Arena综合评价步骤。利用Arena建模软件构建训练流程动态仿真系统并验证有效性,采用单变量数值仿真法、均方差决策法完成了影响因素客观赋权,通过线性加权综合法得出综合评价系数,实现资源配置效率评价与方案优选。实例应用说明了该方法的有效性,同时,该方法可推广至其他类似的训练过程中,亦可为其他多因素、多指标评价决策过程提供借鉴。     

多源文本数据真值发现方法

针对传统真值发现算法无法直接应用于文本数据的问题,提出基于深度神经网络面向多源文本数据的真值发现算法(NN_Truth)。根据文本答案多因素性、词语使用多样性以及文本数据稀疏性等特点,将“数据源-答案”向量作为网络输入,识别答案真值向量作为网络输出,依据真值发现的一般假设,无监督学习各数据源答案向量间关联关系,并最终获得答案真值。实验结果表明,该算法适用于文本数据真值发现场景,较基于检索的方法及传统真值发现算法效果更优。     

高性能互连网络中端口阻塞故障预测方法

随着系统规模、芯片功耗和链路速率的提升,高性能互连网络的整体故障率也不断上升,传统运维方式将难以为继,给高性能计算系统整体可靠性和可用性带来了巨大挑战。针对网络端口阻塞这类严重网络故障,提出无监督算法的预测模型。该模型从历史信息中挖掘征兆性规律并形成新的特征向量,应用K-means聚类算法对特征向量进行学习归类。在预测时,结合端口当前状态,利用二次指数平滑算法对未来状态进行预测,将得到的新特征向量使用K-means算法预判是否会发生阻塞故障。利用拓扑结构信息,分别对叶交换机和根交换机构建预测子模型,进而提升预测的精确率。结果表明,该预测模型能保持在召回率为88.2%的前提下,达到65.2%的准确率,可为运维人员提供有效的辅助。     

Visual-simulation region proposal and generative adversarial network based ground military target recognition

Ground military target recognition plays a crucial role in unmanned equipment and grasping the battlefield dynamics for military applications, but is disturbed by low-resolution and noisy-representation. In this paper, a recognition method, involving a novel visual attention mechanism-based Gabor region proposal sub-network (Gabor RPN) and improved refinement generative adversarial sub-network (GAN), is proposed. Novel central–peripheral rivalry 3D color Gabor filters are proposed to simulate retinal structures and taken as feature extraction convolutional kernels in low-level layer to improve the recognition accuracy and framework training efficiency in Gabor RPN. Improved refinement GAN is used to solve the problem of blurry target classification, involving a generator to directly generate large high-resolution images from small blurry ones and a discriminator to distinguish not only real images vs. fake images but also the class of targets. A special recognition dataset for ground military target, named Ground Military Target Dataset (GMTD), is constructed. Experiments performed on the GMTD dataset effectively demonstrate that our method can achieve better energy-saving and recognition results when low-resolution and noisy-representation targets are involved, thus ensuring this algorithm a good engineering application prospect.     

流体网络在线性能评估与故障诊断方法

在流体网络机理建模的基础上,研究了各种性能退化因素对流体网络特性的影响,通过对网络当前特征参数与理想状态的偏差分析,给出了量化的流体网络性能评价指标,然后以该指标为依据,结合网络的拓扑结构和动态数据,实现了在线性能评估、故障诊断和故障定位。将该方法应用于实际工程案例,并通过仿真结果和实际情况的对比,证明了该方法的实用性和诊断评估结果的准确性。     

A robust sensor covering and communication problem

We consider the problem of placing sensors across some area of interest. The sensors must be placed so that they cover a fixed set of targets in the region, and should be deployed in a manner that allows sensors to communicate with one another. In particular, there exists a measure of communication effectiveness for each sensor pair, which is determined by a concave function of distance between the sensors. Complicating the sensor location problem are uncertainties related to sensor placement, for example, as caused by drifting due to air or water currents to which the sensors may be subjected. Our problem thus seeks to maximize a metric regarding intrasensor communication effectiveness, subject to the condition that all targets must be covered by some sensor, where sensor drift occurs according to a robust (worst‐case) mechanism. We formulate an approximation approach and develop a cutting‐plane algorithm to solve this problem, comparing the effectiveness of two different classes of inequalities. © 2015 Wiley Periodicals, Inc. Naval Research Logistics 62: 582–594, 2015     

基于遗传BP神经网络的装甲装备器材需求预测

装甲装备器材保障具有规模大、时间紧、消耗大、不确定因素多、决策难度大等特点。准确的需求预测是实施主动的、精细化的器材保障的重要前提条件。利用BP神经网络较强自学习能力和自适应能力对器材需求规律进行学习,并借助遗传算法提高BP神经网络的收敛速度,设计了一种基于遗传算法改进的BP神经网络模型预测方法,对装甲装备器材进行需求预测。通过实例计算表明,该方法比单纯BP神经网络方法具有预测精度高、收敛速度快的优点。     

基于粗糙集理论的飞行轨迹识别

针对飞行轨迹人工识别对专家经验依赖性强,识别准确率不稳定的问题,提出一种基于粗糙集理论的飞行轨迹识别方法。首先,对飞行轨迹进行特征提取,获取用于轨迹类别判定的条件属性和决策属性;其次,利用模糊自组织神经网络对飞行轨迹决策表离散化,采用启发式遗传约简算法对条件属性约简,消除了冗余属性对识别结果的影响;最后,利用实际飞参数据进行轨迹识别,结果证明该方法准确有效。