受强弱关系理论启发的改进人工蜂群算法

针对原人工蜂群算法在寻优过程中存在收敛精度不高、容易陷入局部最优的问题,提出一种改进人工蜂群算法(SWT-ABC)。将社会学中强弱关系模型化并引入多子群矩阵式蜂群结构,定义了强关系个体从三个方向随机引导搜索,加快算法收敛速度和提高收敛精度;为增强算法跳出局部最优的能力,定义了弱关系个体交互以实现子群间信息交流来提升种群多样性;增加侦查蜂反向学习机制并确定合适的蜜源上限,能有效提升目标函数评价次数的利用效率。通过基准测试函数的数值实验并与12种改进算法进行对比,改进后的人工蜂群算法收敛精度更高、全局寻优能力更强,并且在高维优化问题求解中仍具备良好的收敛性能。     

Nonstationary filtering of transmission tomograms in high photon counting noise

The algorithm presented here for nonstationary constrained least-squares filtering deals with the transmission photon counting noise problem in the presence of limited dosages and highly nonhomogeneous fields. This technique consists of a design for a set of nonstationary filters, tuned to local noise autocorrelation functions in the reconstructed image. Estimates of noise properties can be taken from object detection information. The approach yields improvement in the tradeoff between noise levels and resolution of the image by exchanging artifacts for less colored noise, which is more easily overcome by the viewer.     

信息化时代Java SE课程网络资源建设与应用研究

运用超星泛雅平台创建了Java SE课程的教学PPT、教学视频、课后作业等网络资源,学生利用所创建Java SE课程的网络资源进行课前预习、课中学习、练习及完成教师发布课后作业等环节的训练.在超星泛雅平台上积累了学生学习Java SE课程网络资源丰富的过程性学习数据,通过分析所积累的学习数据,能有效地评价每一位学生的学习效果和老师应用Java SE课程网络资源的教学效果.     

支持O2O的时空众包激励技术研究

在这个流行元素日新月异的时代,消费者的口味越来越难以把控,商家的判别标准将极大决定店铺的业绩水平.文章介绍了时空众包的概念,研究在商城背景下,利用时空众包思想设计一种能持续有效执行的机制来提高O2O模式的资源利用率问题.     

基于GPS数据和压缩栅格算法的城市路网电子地图快速构建方法

路网电子地图对城市交通管理领域有重要的作用,但是地图服务机构提供的接口功能有限,商用地图一般价格高昂;而一些基于GPS数据的地图生成算法非常复杂、速度慢,在一些实际应用中受到限制,如区域地图即时更新、道路临时管制、应急路线规划等.为此提出一种直观、可扩展、操作灵活的城市道路地图构建方式.通过奥赛罗坐标化的方式,对单位栅格的表示方式进行重新设计,在改善栅格地图的拓扑表达能力、弥补栅格地图存储空间较大和处理时运算资源需求较多等弊端的同时,实现了地图的快速构建.文章以公共交通服务评估和路网韧性评价为例,展示了这一方法的应用潜力.     

Re-GAN：残差生成式对抗网络算法

目的 生成式对抗网络（generative adversarial network,GAN）是一种无监督生成模型,通过生成模型和判别模型的博弈学习生成图像。GAN的生成模型是逐级直接生成图像,下级网络无法得知上级网络学习的特征,以至于生成的图像多样性不够丰富。另外,随着网络层数的增加,参数变多,反向传播变得困难,出现训练不稳定和梯度消失等问题。针对上述问题,基于残差网络（residual network,ResNet）和组标准化（group normalization,GN）,提出了一种残差生成式对抗网络（residual generative adversarial networks,Re-GAN）。方法 Re-GAN在生成模型中构建深度残差网络模块,通过跳连接的方式融合上级网络学习的特征,增强生成图像的多样性和质量,改善反向传播过程,增强生成式对抗网络的训练稳定性,缓解梯度消失。随后采用组标准化（GN）来适应不同批次的学习,使训练过程更加稳定。结果 在Cifar10、CelebA和LSUN数据集上对算法的性能进行测试。Re-GAN的IS （inception score）均值在批次为64时,比DCGAN （deep convolutional GAN）和WGAN （Wasserstein-GAN）分别提高了5%和30%,在批次为4时,比DCGAN和WGAN分别提高了0.2%和13%,表明无论批次大小,Re-GAN生成图像具有很好的多样性。Re-GAN的FID （Fréchet inception distance）在批次为64时比DCGAN和WGAN分别降低了18%和11%,在批次为4时比DCGAN和WGAN分别降低了4%和10%,表明Re-GAN生成图像的质量更好。同时,Re-GAN缓解了训练过程中出现的训练不稳定和梯度消失等问题。结论 实验结果表明,在图像生成方面,Re-GAN的生成图像质量高、多样性强;在网络训练方面,Re-GAN在不同批次下的训练具有更好的兼容性,使训练过程更加稳定,梯度消失得到缓解。     

配电终端的嵌入式程序更新系统设计

针对配电网中配电终端数量多、分布广以及经常需要维护更新的问题,为提高现场程序更新的工作效率,设计实现了一种配电终端嵌入式程序更新的系统.本系统以配电终端Flash的扇区大小为单位进行读写,既减少了程序对Flash的操作次数,也避免了 一帧数据跨扇区存储的问题,使得整个程序更新过程有效减少了对Flash的读写操作,节省了时间.本文介绍了具体实现策略,并已在部门产品中得到了应用和验证.     

A pair of parallel differential magnetic‐field probes with high measurement accuracy and high electric‐field suppression ratio

Magnetic‐field probes can be used for electromagnetic interference measurement of high‐speed circuits. The main magnetic probe performance includes sensitivity, spatial resolution, electric‐field suppression ratio (EFSR), and measurement accuracy. In this article, a pair of differential magnetic‐field probes is proposed to improve measurement accuracy without reducing sensitivity. The proposed differential probes consist of two asymmetric loop probes, which are designed in the same plane and separated by a row of periodic vias. The proposed differential probes are fabricated under PCB process. High accuracy can be achieved by measuring difference between outputs of the two probes. In addition, EFSR can be improved by size optimization of the differential magnetic‐field probes. Simulation and measurement results show the operating bandwidth is from 100 MHz to 12 GHz, the measurement error is 3.4% and the EFSR is about 40 dB. The proposed probes have higher measurement accuracy and higher EFSR than the conventional single probe, and larger operation bandwidth than the stacked differential probes.     

Supervised discrete hashing through similarity learning

Multimedia Tools and Applications - Supervised hashing has achieved better accuracy than unsupervised hashing in many practical applications owing to its use of semantic label information. However,...     

An SVD-based adaptive robust speech steganography using MDCT coefficient

Multimedia Tools and Applications - Speech is one of the essential ways of communication. The study of speech steganography provides great value in information security. To improve imperceptibility...