基于联合神经网络学习的中文电力计量命名实体识别

随着电力计量业务的不断扩展，迫切需要由业务信息、技术知识、行业标准及其内在联系所组成的电力计量知识图谱，为电网的决策和发展提供更为全面有效的支持。命名实体识别是构建知识图谱的基础。针对电力计量领域需要，结合中文分词技术特点，基于联合学习思想，提出了一种基于联合学习的中文电力计量命名实体识别技术。该技术联合CNN-BLSTM-CRF模型与整合词典知识的分词模型，使其共享实体类别和置信度；同时将2个模型的先后计算顺序改为并行计算，减少了识别误差累积。结果表明，在不需要人工构建特征的情况下，方法的正确率、召回率、F值等均显著优于以往方法。     

基于视觉辐条图案识别方法的蓝宝石引晶机制研究

工业生产蓝宝石晶体过程中,引晶步骤有着至关重要的地位。引晶必须在温度梯度较小,温度分布趋于稳定的条件下进行。目前,工业生产蓝宝石主要依靠人工经验操控籽晶杆实现引晶操作,但是人工引晶操作的准确性不高会导致成品品质不佳、资源浪费。为此,本文提出一种基于蓝宝石视觉辐条图案识别方法来检测蓝宝石熔体状态自由液面状态,从而实现一种高效率引晶的机制。此方法利用经典骨架化算法细化辐条图案,Harris算子实现特征信息的提取,提取的特征信息放入运动轨迹模型中判断熔体稳定性,分析液面温度分布稳定性从而实现引晶。结果表明,此算法具有有效性,蓝宝石晶体引晶效率大大提高,生产出的成品良率也有提升,可有效指导蓝宝石的工业生产。     

High-fidelity Recognition of Organotrifluoroborate Anions (R−BF3−) as Designer Guest Molecules

The recognition of boron compounds is well developed as boronic acids but untapped as organotrifluoroborate anions (R−BF₃⁻). We are exploring the development of these and other designer anions as anion-recognition motifs by considering them as substituted versions of the parent inorganic ion. To this end, we demonstrate strong and reliable binding of organic trifluoroborates, R−BF₃⁻, by cyanostar macrocycles that are size-complementary to the inorganic BF₄⁻ progenitors. We find that recognition is modulated by the substituent's sterics and that the affinities are retained using the common K⁺ salts of R−BF₃⁻ anions.     

Bidentate Boron Lewis Acids: Selectivity in Host–Guest Complex Formation

Bidentate boron Lewis acids based on 1,8‐diethynylanthracene were synthesised in two steps by initial stannylation of the terminal alkynes and subsequent tin–boron exchange with different chloroboranes. The reactions were very selective, and the target compounds were obtained in high purity and good to excellent yields. Complexation experiments of 1,8‐bis[(diphenylboranyl)ethynyl]anthracene with nitrogen bases (pyridine, pyrimidine, TMEDA) afforded three stable adducts, which were structurally characterised by X‐ray diffraction. Competition experiments demonstrated the selective exchange of guests, and quantum‐chemical calculations provided information on their energetics. NMR experiments at low temperature gave insight into the dynamic behaviour of the TMEDA adduct.     

Preparation of aptamer‐bound polyamine affinity monolithic column via a facile triazine‐bridged strategy and application to on‐column specific discrimination of ochratoxin A

Developing a high‐performance modification protocol is critical for efficiently fabricating affinity monolith. Herein, by using 2,4,6‐trichloro‐1,3,5‐triazine as the linker, a simple triazine‐bridged approach was proposed for efficiently fabricating aptamer‐grafted polyhedral oligomeric silsesquioxane‐polyethyleneimine affinity monolith with high specificity toward ochratoxin A. Experimental parameters, column characteristics and specificity performances of the resultant affinity monolith were investigated in detail. Under the optimal conditions, 2,4,6‐trichloro‐1,3,5‐triazine was rapidly grafted on the polyamine matrix, and effectively applied to the subsequent bridge linkage of aptamers. It was simple and effective, which resulted in a significant decrease of modification time, excellent properties including the high coverage density of aptamer up to 1799 pmol/μL and sensitive detection of ochratoxin A as low as 10 pg/mL in beer samples. This protocol provides a facile approach for fabricating aptamer‐grafted polyamine affinity monoliths with highly selective discrimination performance.     

Recent development trends for chiral stationary phases based on chitosan derivatives,cyclofructan derivatives and chiral porous materials in high performance liquid chromatography

The separation of enantiomers by chromatographic methods, such as gas chromatography, high‐performance liquid chromatography and capillary electrochromatography, has become an increasingly significant challenge over the past few decades due to the demand of pharmaceutical, agrochemical, and food analysis. Among these chromatographic resolution methods, high‐performance liquid chromatography based on chiral stationary phases has become the most popular and effective method used for the analytical and preparative separation of optically active compounds. This review mainly focuses on the recent development trends for novel chiral stationary phases based on chitosan derivatives, cyclofructan derivatives, and chiral porous materials that include metal‐organic frameworks and covalent organic frameworks in high‐performance liquid chromatography. The enantioseparation performance and chiral recognition mechanisms of these newly developed chiral selectors toward enantiomers are discussed in detail.     

Tunable Orthogonal Reversible Covalent (TORC) Bonds: Dynamic Chemical Control over Molecular Assembly

Dynamic assembly of macromolecules in biological systems is one of the fundamental processes that facilitates life. Although such assembly most commonly uses noncovalent interactions, a set of dynamic reactions involving reversible covalent bonding is actively being exploited for the design of functional materials, bottom‐up assembly, and molecular machines. This Minireview highlights recent implementations and advancements in the area of tunable orthogonal reversible covalent (TORC) bonds for these purposes, and provides an outlook for their expansion, including the development of synthetically encoded polynucleotide mimics.     

基于双向循环神经网络的汉语语音识别*

当前基于深度神经网络模型中,虽然其隐含层可设置多层,对复杂问题适应能力强,但每层之间的节点连接是相互独立的,这种结构特性导致了在语音序列中无法利用上下文相关信息来提高识别效果,而传统的循环神经网络虽然做出了改进,但是只能对上文信息进行利用。针对以上问题,该文采用可以同时利用语音序列中上下文相关信息的双向循环神经网络模型与深度神经网络模型相结合,并应用于语音识别。构建具有5层隐含层的模型,其中第3层为双向循环神经网络结构,其他层采用深度神经网络结构。实验结果表明:加入了双向循环神经网络结构的模型与其他模型相比,较好地提高了识别正确率;噪声对双向循环神经网络汉语识别有重要影响,尤其是训练集和测试集附加噪声类型不同时,单一的含噪声语音的训练模型无法适应不同噪声类型的语音识别;调整神经网络模型中隐含层神经元数量后,识别正确率并不是一直随着隐含层中神经元数量的增加而增加,神经元数量数目增加到一定程度后正确率出现了降低的趋势。     

A DNA Switch for Detecting Single Nucleotide Polymorphism within a Long DNA Sequence Under Denaturing Conditions

DNA detection is usually conducted under nondenaturing conditions to favor the formation of Watson–Crick base-paring interactions. However, although such a setting is excellent for distinguishing a single-nucleotide polymorphism (SNP) within short DNA sequences (15–25 nucleotides), it does not offer a good solution to SNP detection within much longer sequences. Here we report on a new detection method capable of detecting SNP in a DNA sequence containing 35–90 nucleotides. This is achieved through incorporating into the recognition DNA sequence a previously discovered DNA molecule that forms a stable G-quadruplex in the presence of 7 molar urea, a known condition for denaturing DNA structures. The systems are configured to produce both colorimetric and fluorescent signals upon target binding.     

贝叶斯优化卷积神经网络公共场所异常声识别*

针对公共场所异常声的感知和识别问题,提出一种基于贝叶斯优化卷积神经网络的识别方法。提取声信号的Gammatone倒谱系数、倍频程功率谱、短时能量和谱质心,组合成声信号的特征图。构建卷积神经网络作为分类器,利用递增的卷积核设置和池化操作处理不同尺度的特征。基于贝叶斯优化算法优化卷积神经网络的模型参数,对包括火苗噼啪声、婴儿啼哭声、烟花燃放声、玻璃破碎声和警报声的5种公共场所异常声进行识别。该方法的识别结果与基于不同的特征提取和分类器方案得到的识别结果进行比较,结果表明该方法的识别效果优于其他特征提取和分类器方案的识别效果。最后分析了该方法在不同信噪比噪声干扰下的识别结果,验证了该方法的有效性。