基于NIR高光谱成像技术的长枣虫眼无损检测

为了研究快速识别虫眼枣与正常枣的有效方法,利用特征波长主成分分析法结合波段比算法进行虫眼枣识别。首先,利用NIR高光谱成像系统采集130个长枣(50个正常、80个虫眼枣)图像,提取并分析不同类型长枣特征区域的平均光谱曲线,对970~1 670 nm范围内的光谱数据进行主成分分析,确定7个特征波长(990,1 028,1 109,1 160,1 231,1 285,1 464 nm)。然后,对长枣图像做主成分分析,选择PC2图像进行虫眼识别,虫眼与正常枣的识别率分别为67.5%、100%。为了进一步提高虫眼枣的识别率,采用波段比(R1231/R1109)对未识别的虫眼枣进行再次识别,识别率提高到90%。结果表明,基于NIR高光谱成像技术的检测方法对虫眼枣识别是可行的,同时也为多光谱成像技术应用于在线检测长枣品质提供了理论依据。     

基于卷积神经网络的水下湍流探测技术

针对水下湍流的复杂性和多变性对水下航行器性能和姿态控制产生的挑战,提出使用卷积神经网络来测量水下湍流的温差耗散率XT。首先,采用功率谱反演法和惠更斯-菲涅尔原理仿真生成了受水下湍流影响的散斑图像数据集。随后,利用卷积神经网络提取这些受湍流影响的散斑图像中的特征信息,并对温差耗散率XT进行估计。最后,通过现场实验数据集验证了所提出方法的可行性。实验结果表明,所提出的神经网络在实地实验数据集和模拟仿真数据集上表现出相似的分类精度和损失曲线,其测量准确率分别为98.8%和99.2%。这一研究为水下环境监测和资源勘探领域提供了重要的参考,对于光学图像处理和湍流研究等相关领域具有实际意义。     

声速测量实验原理讨论

本文针对声速测量实验中接收端声压变化的理论推导还不够完善,尤其是对相位比较法的原理解释不清楚的问题,从振幅与声压的关系出发,建立了基于声压方程的多次反射模型.该模型利用平面简谐波函数、声压反射系数公式和波的叠加原理,推导出了接收端的声压振幅和相位随距离的变化公式,并利用该公式解释了驻波法和相位比较法测量的原理.仿真结果表明接收端的声压振幅和相位并不是严格的周期函数,其测量的精度与声吸收系数和声压反射系数等因素有关.最后进行了实验验证,实验测量结果与本文模型的计算结果吻合较好.本文的研究为正确理解该实验的物理本质提供理论指导.     

Adaptive interaction driven by the learning effect in the spatial prisoner's dilemma

We propose a computing model in which individuals can automatically adjust their interaction intensity with their mentor according to the learning effect. This model is designed to investigate the cooperative dynamics of the spatial prisoner’s dilemma. More specifically, when the cumulative payoff of a learner is more than his reference earning, he will strengthen his interaction with his mentor; otherwise, he will reduce it. The experimental results indicate that this mechanism can improve the ...     

1.3 GHz超导腔磁通排出效应研究

设计并搭建了一套高精度的磁场测量和补偿系统,并结合中国科学院高能物理研究所（IHEP）的2K超导腔垂直测试平台对1.3 GHz 单加速间隙超导腔的磁通排出效应开展了实验研究：利用研制的磁场测量和补偿系统能够精密地测量超导腔赤道位置磁场,并能够将磁场补偿至小于5.0×10⁻⁸ T;并对超导腔不同表面温度梯度下的磁通排出效应进行了测量分析;对钉扎了磁场的超导腔进行了射频性能测试,研究了超导腔电阻对磁通钉扎的敏感度,以及在不同电场梯度下超导腔的表面电阻变化情况。结果表明,研制的高精度磁场测量和补偿系统能够满足超导腔磁通排出研究的需求;高的超导腔表面温度梯度有利于磁通的排出;磁通钉扎电阻的敏感度随着加速电场梯度的增加而增大,导致超导腔的性能下降。此实验研究也为后续超导腔的研制奠定了一定基础。     

基于EKF的水下LD通信精对准控制算法

高速激光通信中接收机与光斑中心很难处于精对准状态,导致水下光通信链路难以稳定建立.首先采用蒙特卡洛仿真统计法分析激光光子在海水中传输的接收光强分布规律,再通过实验对接收端的光斑图像进行采样分析,利用曲线拟合得到接收器位置与接收光强的关系.仿真与实验结果表明:光束经过25 m的水下传输,接收光强分布仍近似为高斯分布.采用非线性估计算法(扩展卡尔曼滤波)与基本状态控制反馈理论,根据接收光强度估计接收器当前位置与最大光强处的距离,通过反馈算法实现接收端与光斑中心的主动跟踪对准.算法仿真结果显示,接收端对准误差在2 mm以下,稳定后接收效率超过98%.     

Improvement of high-frequency properties of Co_2FeSi Heusler films by ultrathin Ru underlayer

Heusler Co2FeSi films with a uniaxial magnetic anisotropy and high ferromagnetic resonance frequency fr were deposited by an oblique sputtering technique on Ru underlayers with various thicknesses tRufrom 0 nm to 5 nm.It is revealed that the Ru underlayers reduce the grain size of Co2FeSi,dramatically enhance the magnetic anisotropy field HK induced by the internal stress from 242 Oe(1 Oe=79.5775 A·m^-1)to 582 Oe with an increment ratio of 2.4,while a low damping coefficient remains.The result of damping implies that the continuous interface between Ru and Co2FeSi induces a large in-plane anisotropic field without introducing additional external damping.As a result,excellent high-frequency soft magnetic properties with fr up to 6.69 GHz are achieved.     

读者来信

以下刊登的几篇读者来信，对我刊一些文章中的错误进行批评指正．这些来信反映了广大读者对我刊的爱护和支持，今后我们将注意改进工作．     

避免假重传使TCP更健壮的RR-CETEN算法

许多研究表明,在高并行度和不同链接速率的无线网和卫星网中都存在包的重排序．由于重排序将导致对网络拥塞过高的估计,它大大降低了网络的TCP性能．CETEN（Cumulative Explicit Transport Error Notification）虽然解决了出错丢包对网络拥塞估计偏差的问题,但由于没有考虑无线网络下包的重排序情况,同时也使LEAST（Loss Estimations AlgorithmS for TCP）算法的丢包估计值不准确,导致出错丢包率产生偏差．为此,通过在接收方增加显式的假重传通告方案,使发送方不仅能区分丢失（出错和拥塞）的包,还能处理重排序包,提高了LEAST算法的精确度,使TCP更加健壮．试验仿真结果表明这种方法的有效性．     

All polymer asymmetric Mach-Zehnder interferometer waveguide sensor by imprinting bonding and laser polishing

We present an all polymer asymmetric Mach–Zehnder interferometer(AMZI) waveguide sensor based on imprinting bonding and laser polishing method. The fabrication methods are compatible with high accuracy waveguide sensing structure. The rectangle waveguide structure of this sensor has three sensing surfaces contacting the test media, and its sensing accuracy can be increased 5 times compared with that of one surface sensing structure. An AMZI device structure is designed. The single mode condition, the length of the sensing arm, and the length deviation between the sensing arm and the reference arm are optimized. The length deviation is optimized to be 19.8 μm in a refractive index range between1.470 and 1.545. We fabricate the AMZI waveguide by lithography and wet etching method. The imprinting bonding and laser polishing method is proposed and investigated. The insertion loss is between-80.36 dB and-10.63 dB. The average and linear sensitivity are 768.1 d B/RIU and 548.95 dB/RIU, respectively. And the average and linear detection resolution of the sensor are 1.30×10~(-6) RIU(RIU: refractive index unit) and 1.82×10~(-5) RIU, respectively. This sensor has a fast and cost-effective fabrication process which can be used in the cases of requiring portability and disposability.