Passivating Defects of Perovskite Solar Cells with Functional Donor-Acceptor–Donor Type Hole Transporting Materials

In this study, a series of donor–acceptor–donor (D-A-D) type small molecules based on the fluorene and diphenylethenyl enamine units, which are distinguished by different acceptors, as holetransporting materials (HTMs) for perovskite solar cells is presented. The incorporation of the malononitrile acceptor units is found to be beneficial for not only carrier transportation but also defects passivation via Pb–N interactions. The highest power conversion efficiency of over 22% is achieved on cells based on V1359, which is higher than that of spiro-OMeTAD under identical conditions. This st shows that HTMs prepared via simplified synthetic routes are not only a low-cost alternative to spiro-OMeTAD but also outperform in efficiency and stability state-of-art materials obtained via expensive cross-coupling methods.     

MAML2: meta reinforcement learning via meta-learning for task categories

Meta-learning has been widely applied to solving few-shot reinforcement learning problems, where we hope to obtain an agent that can learn quickly in a new task. However, these algorithms often ignore some isolated tasks in pursuit of the average performance, which may result in negative adaptation in these isolated tasks, and they usually need sufficient learning in a stationary task distribution. In this paper, our algorithm presents a hierarchical framework of double meta-learning, and the whole framework includes classification, meta-learning, and re-adaptation. Firstly, in the classification process, we classify tasks into several task subsets, considered as some categories of tasks, by learned parameters of each task, which can separate out some isolated tasks thereafter. Secondly, in the meta-learning process, we learn category parameters in all subsets via meta-learning. Simultaneously, based on the gradient of each category parameter in each subset, we use meta-learning again to learn a new meta-parameter related to the whole task set, which can be used as an initial parameter for the new task. Finally, in the re-adaption process, we adapt the parameter of the new task with two steps, by the meta-parameter and the appropriate category parameter successively. Experimentally, we demonstrate our algorithm prevents the agent from negative adaptation without losing the average performance for the whole task set. Additionally, our algorithm presents a more rapid adaptation process within re-adaptation. Moreover, we show the good performance of our algorithm with fewer samples as the agent is exposed to an online meta-learning setting.     

多应用场景低压配网无缝合环转电解决方案

合环转电作为一种重要的网络重构技术,对提高配电网运行的可靠性与经济性具有重要意义。常规合环转电不仅存在合环冲击,而且合环点的选取受到两侧电源相角差的限制。文章针对广州配电网存在的大角度差不停电转供技术难题,提出两种无缝合环转电解决方案,分别是基于串并联补偿的合环方案和基于背靠背拓扑的合环方案。它们能够适用于多种转供场景,并提供交、直流两种转供电源接口。文章给出了每种方案的接线方式,对它们的结构特点、性能参数以及工作原理进行了分析。为了适应配电网三相不平衡与谐波污染的运行环境,从结构与控制两方面提出了装置的改进措施。最后,仿真验证了几种典型场景下无缝合环转电装置的运行原理,从而为实际开展低压无缝合环转电提供计算支撑。     

BP神经网络结合变量选择方法在牛奶蛋白质含量检测中的应用北大核心CSCD

牛奶中的蛋白质含量会影响牛奶的品质,利用高光谱图像的光谱特征信息研究对牛奶蛋白质含量预测的可行性。本文提出一种基于竞争性自适应重加权算法(competitive adaptive reweighted sampling, CARS)和连续投影算法(successive projections algorithm, SPA)结合多层前馈神经网络(back propagation, BP)的预测建模方法,实验以含有不同浓度蛋白质的牛奶为对象,利用可见光/近红外高光谱成像系统共采集到5种牛奶共计250组高光谱数据,通过实验对比选择采用标准化方法对获取到的吸收光谱预处理,然后采用CARS结合SPA筛选特征波长,得到18个特征波长,建立CARS-SPA-BP模型,经过试验,CARS-SPA-BP模型的训练集决定系数和测试集决定系数R;和R;分别达到0.971和0.968,训练集均方根误差(root mean square error of calibration,RMSEC)和测试集均方根误差(root mean square error of prediction,RMSEP)达到了0.033和0.034。研究发现,采用CARS结合SPA筛选的牛奶特征波长建立的多层前馈神经网络模型,其模型预测结果与全波长建模相比并没有明显降低,因此将CARS结合SPA用于波长筛选并且结合BP神经网络基本可以完成对牛奶蛋白质含量的预测。为验证CARS-SPA-BP模型的预测能力,在相同数据环境下,使用较为传统的偏最小二乘回归(partial least squares regression, PLSR)进行建模,实验结果表明,CARS-SPA-BP相较于PLSR,R;和RMSEP均有明显提升。研究表明,CARS-SPA-BP可充分利用牛奶光谱特征信息实现较高精度的牛奶蛋白质含量检测。     

Research on the properties of ZnO1-xSx thin films modified by sulfur doping for CIGS solar cells

ZnO1-xSx thin films modified by sulfur doping were prepared on glass substrates by chemical bath deposition (CBD) for studying the effect of thiourea concentration on the thin film properties. The obtained ZnO1-xSx thin films were characterized by scanning electron microscopy (SEM), which shows the surfaces of ZnO1-xSx thin films deposited under the thiourea concentration of 0.14 M are more compact. X-ray diffraction (XRD) measurement shows that the ZnO1-xSx thin films with hexagonal crystal structure had strong diffraction peaks and better crystallinity. The optical transmittance of the ZnO1-xSx thin films with 0.14 M thiourea concentration is above 80% in the wavelength range of 300—900 nm. According to the measurement results from spectrophotometer, the ZnO1-xSx band gap energy value Eg varies nonlinearly with different S/(S+O) ratio x, and increases with the increase of x. There is a band gap value of 2.97 eV in the ZnO1-xSx thin films deposited under 0.14 M thiourea concentration. Therefore, the thin films have better structural, optical and electric properties, and are more suitable for the buffer layers of copper indium gallium selenide (CIGS) thin film solar cells.     

配电网电缆故障行波衰减特性及定位方法研究

行波衰减特性和不同的故障条件是影响在线行波定位精度的重要因素。本文利用建立了10kV配电网电缆行波传播衰减模型,分别研究了行波频率参数、时域参数与传输距离的关系。最后分析了不同因素对行波定位结果的影响。结果表明,行波幅值比、波头时比、波尾时比和脉宽比参数与传输距离呈二次函数关系。其中幅值比随传输距离单调减小,其他三个参数随传输距离单调增大。此外,故障行波头尾时间和故障点位置对故障定位结果有一定影响。如果初始波头尾时间短,传输距离长,会导致定位误差增大。研究发现故障过渡电阻对定位结果影响不大。最后,本文基于行波传播衰减特性的研究,提出了基于衰减特性的行波故障定位方法。     

一种低轨双星高脉冲重复频率雷达信号的定位模糊消除算法

针对低轨(LEO)双星到达时间差(TDOA)定位系统中雷达信号的定位模糊问题,该文分析了高重频信号的模糊特性,演示了虚假定位点的产生过程,提出一种基于时差序列变化趋势匹配(VTMT)的定位模糊消除算法.在测量时差序列与定位结果对应的理论时差序列之间的残差近似服从正态分布的条件下,通过两个序列的欧式距离作为相似程度衡量依据,筛选出测量时差线与理论时差线最为接近的定位点作为目标辐射源真实位置.校验台测试信号验证了该算法的有效性.仿真结果表明,该算法能够显著提高无模糊定位概率,对长基线时差定位系统中雷达信号的定位具有较高的应用价值.