The recognition and localization of insulators adopting SURF and IFS based on correlation coefficient

The recognition and localization of insulators is a crucial process of insulators’ failure detection. Aiming to identify insulators in test aerial image under the complicated background with no template and segmentation, a novel method of the recognition and localization of insulators based on SURF (Speeded Up Robust Features) and IFS (Intuitionistic Fuzzy Set) based on correlation coefficient is proposed as following: SURF is used to extract a set of feature points of the test aerial image. Then use IFS based on correlation coefficient to partition points into k classes. Find all the connected regions for each class, and calculate minimum circumscribed rectangle and shape characteristic value for each region to identify insulators, and localize them by rectangular boxes. Thus we can detect all the insulators from test aerial image correctly. Experimental results show that the proposed approach can obtain high precision in the recognition and localization of insulators.     

对称零面积变换结合L-M拟合自动识别重叠光谱峰

Raman光谱分析中,由于仪器光谱分辨率的限制和复杂的目标成分,经常存在着谱峰重叠现象。容易导致谱峰参数提取乃至样品成分分析错误,更为现场光谱的自动分析处理增加了难度。重叠光谱峰的识别已成为现场光谱学处理的难点,针对这个问题,建立了一套自动识别重叠谱峰的方法：以多个高斯峰的组合作为重叠光谱解析模型,首先以对称零面积变换寻峰方法确定高斯峰的个数和初步的峰位、峰高、宽度参数,并以Levenberg-Marquardt方法对获得的初始高斯峰参数进行拟合优化,最终获得各个独立谱峰的拟合参数。对该方法进行了测试,其中以仿真数据进行的算法实验证明,对称零面积变换寻峰获得的初始参数与真实值较为接近,在此基础上进行的参数优化收敛速度快,对峰参数的提取可获得较高的准确度;以具有不同信噪比的实测Raman光谱数据进行的算法实验证明,该方法可以适用于信噪比在较大范围变化的信号,但对信噪比过低的实测信号则容易产生虚假峰和漏峰。研究证明,以对称零面积变换寻峰结合L-M拟合自动识别重叠光谱峰的方法具有一定的实用价值。     

SE-MCNN-CTC的中文语音识别声学模型

针对国内外缺少对振动轮噪声预估的问题,以某型振动轮为研究对象,首先基于动力学有限元理论对振动轮进行频率响应分析,其次采用声学边界元技术对振动轮辐射噪声进行了数值模拟,并通过实验验证了仿真结果的准确性,然后比较了垂直振动与圆周振动两种不同激振形式对辐射噪声的影响,得出垂直振动辐射噪声低的结论,最后对驾驶室声腔模态进行了仿真,与振动轮激振频率相近发生共振。通过调整激振频率,降低了司机耳旁噪声。所得研究成果可为振动轮辐射噪声的预估与改进提供一种切实可行的参考依据。     

基于SVM回归模型的混合气体组分种类光谱识别方法

针对混合气体红外光谱分析中无法采用同一模型同时进行混合气体组分浓度的定量分析和组分种类的定性分析的问题,本文提出了基于SVM回归模型的混合气体组分种类光谱识别方法.通过详细推导,证明混合气体组分种类识别完全可以通过组分浓度分析的SVM回归模型来求解,混合气体组分种类识别是一种特殊的回归.实验结果显示,该方法的混合气体组分种类的正确识别率不小于92.5%.     

茜素应用于识别硼酸的研究

应用紫外-可见分光光度法研究了茜素与硼酸之间的识别行为,结果表明,在pH 9.0Na2HPO4-NaH2PO4缓冲介质中,茜素与硼酸形成1∶1稳定络合物,使茜素的吸收光谱发生明显蓝移,溶液颜色由玫红色变为橙黄色,阴离子F-、Cl-、Br-、I-、Ac-、NO-3、C2O24-等均未引起茜素的吸收光谱及溶液颜色发生明显变化,因此茜素可以选择性识别硼酸。初步探讨了茜素与硼酸之间的相互作用机理,并对茜素识别硼酸与茜素识别阴离子的机理进行了比较。     

卟啉及其配合物在分析化学中应用进展

综述了卟啉及配合物在分析化学中的应用研究进展。重点介绍了其在光度分析、电化学、分子识别、探针等方面的应用。     

新型混合脂质体的制备及其与药物的分子识别研究

设计、合成了含核酸碱基的双亲化合物N6-十四酰基-9-辛基腺嘌呤三甲基溴化铵;用目标化合物和卵磷脂制备混合脂质体受体。分别研究了受体与巴比妥酸、核黄素药物在混合脂质体亲脂区通过互补氢键进行的分子识别。紫外可见光谱研究表明,底物巴比妥酸在258nm的吸收峰,核黄素在222nm,264nm的吸收峰随时间的延长而逐渐降低,这种减色作用归因于脂质体内的巴比妥酸和核黄素在亲脂区分别与双亲化合物上的腺嘌呤形成了氢键。     

The Story of the Little Blue Box: A Tribute to Siegfried Hünig

The tetracationic cyclophane, cyclobis(paraquat-p-phenylene), also known as the little blue box, constitutes a modular receptor that has facilitated the discovery of many host–guest complexes and mechanically interlocked molecules during the past 35 years. Its versatility in binding small π-donors in its tetracationic state, as well as forming trisradical tricationic complexes with viologen radical cations in its doubly reduced bisradical dicationic state, renders it valuable for the construction of various stimuli-responsive materials. Since the first reports in 1988, the little blue box has been featured in over 500 publications in the literature. All this research activity would not have been possible without the seminal contributions carried out by Siegfried Hünig, who not only pioneered the syntheses of viologen-containing cyclophanes, but also revealed their rich redox chemistry in addition to their ability to undergo intramolecular π-dimerization. This Review describes how his pioneering research led to the design and synthesis of the little blue box, and how this redox-active host evolved into the key component of molecular shuttles, switches, and machines.     

Novel recognition mechanisms in biological adhesion

Molecular recognition is generally thought of in terms of bond formation between unique sites on host and complementary guest molecules, but recent studies have revealed new complexities in biological recognition at cell. Adhesion between cell surfaces similarly involves specific interactions between analogous ligands and receptors. Recent force measurements, however, suggest that cell adhesion proteins may bind via multiple interaction sites that can form in a sequential manner. Other studies further show that in some instances, the principal recognition event may not be receptor–ligand docking, but the assembly of a complex pattern of many receptors and ligands.     

Molecular Recognition of Carbonyl Compounds by Uranyl-salophen Based Neutral Receptors Driven by Van Der Waals Forces

The complexation of the salophen-uranyl metallocleft 2 and of its half-cleft analogue 3 with enones and other carbonyl compounds was assessed in chloroform by UV-Vis titration and, occasionally, by FT-IR measurements. Complexes with receptors 2 and 3 are in all cases more stable than those with the control unsubstituted uranyl-salophen 1 , showing that in addition to the primary binding force provided by coordination of the carbonyl oxygen to the uranium, a significant driving force for complexation, typically in the range of 2-3 kcal/mol, results from van der Waals interactions of the guest with the aromatic walls. Replacement of the phenyl group in 3 with larger aromatic residues to give 4 and 5 , led to enhanced complex stabilities, due to more extended contact surfaces between host and guest.