紫色球杆菌视紫红质光谱特性的机器学习研究

近年来，机器学习等人工智能技术被应用于蛋白质工程，其在蛋白质结构、功能预测、催化活性等研究中具有独特优势。在未知蛋白质结构的情况下，将蛋白质序列和功能特性与机器学习相结合，基于序列-活性关系（innovative sequence-activity relationship，ISAR）算法，将蛋白质氨基酸序列数字化，用快速傅里叶变换（fast four transform，FFT）进行预处理，再进行偏最小二乘回归建模，可在数据集较少情况下拟合得到最佳模型。通过机器学习对紫色球杆菌视紫红质（gloeobacter violaceus rhodopsin，GR）的突变体蛋白质氨基酸序列与光谱最大吸收波长进行建模，获得了最佳模型。用最佳索引LEVM760106建模得到的确定系数R² 为0.944，均方误差E为11.64。用小波变换进行的预处理，其R² 虽也约为0.944，但E大于11.64，不及FFT进行的预处理。方法较好地解决了蛋白质序列与功能特性之间的数学建模问题，在蛋白质工程中可为预测更优的突变体提供支持。     

Sentinel-2卫星的多光谱轻量级船舶目标检测算法

近年来深度卷积神经网络在可见光船舶检测方面取得了显著的进展，然而，大多数相关研究是通过改进大型的网络结构来提高检测性能，因此加大了对更高计算机性能的需求。此外，可见光图像难以在云、雾、海杂波、黑夜等复杂场景检测到船舶。针对以上问题，提出了一种融合红（red, R）、绿（green, G）、蓝（blue, B）和近红外（NIR）4个波段光谱信息的由粗到精细的轻量型船舶检测算法。与现有的方法中根据光谱特性利用水体检测算法提取水体区域不同之处是该算法是利用改进的水体检测算法来提取船舶候选区域。为获取更准确的候选区域，对船舶、厚云、薄云、平静海面、杂波海面5种场景中4个波段的像素值进行了统计分析，选取近红外大于阈值作为辅助判断，并以其中心点获取候选区域32×32大小的切片，并对切片进行非极大值抑制，由此获得了船舶粗检测结果。随后构建了轻量级LSGFNet网络对船舶候选区域切片进行精细识别。构建的网络融合了1×1卷积提取的波谱特征与3×3的提取几何特征，为防止光谱特征与几何特征的信息在融合时“信息不流通”，在LSGFNet网络中引入了ShuffleNet中的通道打乱机制，并减小了模型结构，与典型的轻量级网络相比具有更好的效果且模型较小。最后，利用Sentinel-2卫星多光谱10 m分辨率数据构建了512×512大小的1 120组数据进行粗检测，以及32×32大小的6 014组数据进行精细网络训练，其中候选区域粗提取的查全率为98.99%，精细识别网络精确度为96.04%，不同场景下的平均精确度为92.98%。实验表明该算法在抑制云层、海浪杂波等干扰的复杂背景下具有较高的检测效率，且训练时间短、计算机性能需求低。     

Strategy to mitigate single event upset in 14-nm CMOS bulk FinFET technology

Three-dimensional (3D) TCAD simulations demonstrate that reducing the distance between the well boundary and N-channel metal-oxide semiconductor (NMOS) transistor or P-channel metal-oxide semiconductor (PMOS) transistor can mitigate the cross section of single event upset (SEU) in 14-nm complementary metal-oxide semiconductor (CMOS) bulk FinFET technology. The competition of charge collection between well boundary and sensitive nodes, the enhanced restoring currents and the change of bipolar effect are responsible for the decrease of SEU cross section. Unlike dual-interlock cell (DICE) design, this approach is more effective under heavy ion irradiation of higher LET, in the presence of enough taps to ensure the rapid recovery of well potential. Besides, the feasibility of this method and its effectiveness with feature size scaling down are discussed.     

Analysis on First-stage Ignition of n-Heptane at Low Temperatures with a Lumped Skeletal Mechanism

Two-stage ignition exists in the low-temperature combustion process of n-heptane and the first-stage ignition also shows a negative temperature coefficient(NTC) phenomenon. To study key reactions and understand chemical principles affecting the first-stage ignition of n-heptane, a lumped skeletal mechanism with 62 species is obtained based on the detailed NUIGMech1.0 mechanism using the directed relation graph method assisted by sensitivity analysis and isomer lumping. The lumped mechanism shows good performance on ignition delay time under wide conditions. The study revealed that the temperature after the first-stage ignition is higher and a larger amount of fuel is consumed at lower initial temperatures. The temperature at the first-stage ignition is relatively insensitive to the initial temperature. Further sensitivity analysis and reaction path analysis carried out based on the lumped mechanism show that the decomposition of RO₂ to produce alkene and HO₂is the most important reaction to inhibit the first-stage ignitions. The chain branching explosion closely related to the first-stage ignition will be terminated when the rate constant for the RO₂ decomposition is larger than that of the isomerization of RO₂ to produce QOOH. The NTC behavior as well as other characteristics of the first-stage ignition can be rationalized from the competition between these two reactions.     

Mg,N掺杂β-Ga2O3光电性质的第一性原理计算

通过基于密度泛函理论的第一性原理计算,研究了Mg单掺杂、N单掺杂和不同浓度的Mg-N共掺杂β-Ga₂O₃的结构性质、电子性质和光学性质,以期获得性能比较优异的p型β-Ga₂O₃材料。建立了五种模型:Mg单掺杂、N单掺杂、1个Mg-N共掺杂、2个Mg-N共掺杂和3个Mg-N共掺杂β-Ga₂O₃。经过计算,3个Mg-N共掺杂β-Ga₂O₃体系的结构最稳定。此外,在5种模型中,3个Mg-N共掺杂β-Ga₂O₃体系的禁带宽度是最小的,并且N 2p和Mg 3s贡献的占据态抑制了氧空位的形成,从而增加了空穴浓度。因此,3个Mg-N共掺杂β-Ga₂O₃体系表现出优异的p型性质。3个Mg-N共掺杂体系的吸收峰出现明显红移,在太阳盲区的光吸收系数较大,这归因于导带Ga 4s、Ga 4p、Mg 3s向价带O 2p、N 2p的带间电子跃迁。本工作将为p型β-Ga₂O₃日盲光电材料的研究和应用提供理论指导。     

A self-powered and sensitive terahertz photodetection based on PdSe2

With the rapid development of terahertz technology, terahertz detectors are expected to play a key role in diverse areas such as homeland security and imaging, materials diagnostics, biology, medical sciences, and communication. Whereas self-powered, rapid response, and room temperature terahertz photodetectors are confronted with huge challenges. Here, we report a novel rapid response and self-powered terahertz photothermoelectronic (PTE) photodetector based on a low-dimensional material: palladium selenide (PdSe₂). An order of magnitude performance enhancement was observed in photodetection based on PdSe₂/graphene heterojunction that resulted from the integration of graphene and enhanced the Seebeck effect. Under 0.1-THz and 0.3-THz irradiations, the device displays a stable and repeatable photoresponse at room temperature without bias. Furthermore, rapid rise (5.0 μs) and decay (5.4 μs) times are recorded under 0.1-THz irradiation. Our results demonstrate the promising prospect of the detector based on PdSe₂ in terms of air-stable, suitable sensitivity and speed, which may have great application in terahertz detection.     

基于抛光基材的热蒸镀铜箔生长高平整单层石墨烯薄膜

在石墨烯的化学气相沉积工艺中,铜箔是决定石墨烯薄膜质量的重要因素。传统铜箔由于制备工艺的限制,存在大量的缺陷,导致石墨烯薄膜的成核密度较高。本工作选用抛光铝板、抛光不锈钢板、微晶玻璃和SiO₂/Si作为基材,用热蒸镀法制备了不同粗糙度的铜箔,并详细讨论了以该系列铜箔生长高平整度石墨烯薄膜的条件及铜箔对石墨烯薄膜品质的影响。实验结果表明,铜箔以(111)取向为主,与基材分离后,表面具有纳米级平整度。在生长石墨烯后,从SiO₂/Si剥离的铜箔成核密度是4种基材中最小的。同时,从SiO₂/Si剥离的铜箔晶体结构变化最不明显,具有良好的结晶性,表面几乎不存在铜晶界缺陷。当压强为3 000 Pa,氢气和甲烷流速分别为300 mL/min和0.5 mL/min时,可以获得约1 mm横向尺寸的石墨烯单晶晶畴。