Thermodynamic Dual Descriptor

A new definition of the dual descriptor,namely,thethermodynamic dual descriptor,is developed within the grand canonicalpotential formalism.This new definition is formulated to describe the same △f_r(r)=β/2C[f^+(r)-f^-(r)](C>0) physical phenomenon as the original definition proposed by Morell,Grand,and Toro-Labbé(J.Phys.Chem.A2005,109,205),which is characterized by a second-order response of the electron density towards an electron flux.To formulate the new definition,we performed two successive partial derivatives of the average electron density,one with respect to the average number of electrons,and the other with respect to the chemical potential of the electron reservoir.When the derivative is expressed in terms of the three-state ensemble model,in the regime of low temperatures up to temperatures of chemical interest,one finds that the thermodynamic dual descriptor can be expressed as?f_T(r)=(β/2)C[f^+(r)-f^-(r)],whereβ=1/k_BT,Cis a global quantity that depends on the temperature and global electronic properties of the molecule(the first ionization potential and the electron affinity),C=1 for systems with zero fractional charge,and C=C_ω>0(albeit very close to zero)for systems with nonzero fractional charge,,and f^+(r)and f^-(r)are the nucleophilic and electrophilic Fukui functions,respectively.The quantity within the square brackets is the original definition of the dual descriptor.As the local terms(the ones containing regioselectivity information)are equal to those of the dual descriptor,?f_T(r)has the same regioselectivity information,multiplied by the global quantity(β/2)C.This implies that the regioselectivity information contained in the original dual descriptor is preserved at all temperatures different from zero,and for any value of C>0.One of the most important features of this new definition is that it avoidsthe undesired Dirac delta behavior observed when the second order partial derivative of the average density is taken with respect to the average number of electrons,using the exact density dependence of the average number of electrons.

Chemical Reactivity Theory (CRT) Study of the Melanoidin M8: Local Conceptual Density Functional Theory Descriptors

This study evaluated a fixed long-range corrected range-separated hybrid (RSH) density functional associated with the Def2TZVP basis set alongside the Solvation Model based on Density (SMD) for the computation of the structure, molecular properties and chemical reactivity of the M8 intermediate melanoidin pigment. The preference of the active sites pertinent to radical, nucleophilic and electrophilic attacks is made through linking them with the electrophilic and nucleophilic Parr functions, Fukui function indices, and condensed Dual Descriptor. This study showed that the MN12SX density functional is the most suitable one for predicting the chemical reactivity of this system.

亲电取代反应中活性位点预测方法的比较

预测发生亲电取代反应的活性位点具有重要的理论和实际意义.目前已提出了许多基于反应物自身电子结构的预测方法.本文选择14个单取代苯和8个双取代苯作为测试集,对14种预测方法的可靠性进行了详细的比较分析.结果表明,福井函数、平均局部离子化能等体现局部电子软度的方法特别适合含有邻对位定位基的单取代苯和双取代苯体系,但对于含有单个间位定位基的体系,这类方法往往预测失败.基于静电效应的预测方法整体表现明显不如体现局部软度的方法,但更适合含有单个间位定位基的体系.对所有体系预测能力最稳健的是双描述符,因此可以作为普适性的预测方法.

基于双模型竞争机制的目标跟踪算法

为解决背景感知相关滤波器存在的特征表达能力不足和模型漂移问题,本文提出了一种基于双模型竞争机制的目标跟踪算法.一方面,本文基于颜色和梯度信息设计了一种简单高效的特征描述子,以实现更鲁棒的目标表观建模.另一方面,本文分别构建初始模型和变化模型作用于目标搜索区域,并根据两者的跟踪响应图置信度来决定跟踪结果 .跟踪过程中,随着双模型主导地位不断地动态切换,变化模型也被赋予了可逆向学习的能力,从而达到缓解模型漂移的效果.实验结果表明,相比于基准算法,本文算法在OTB2015、TinyTLP和UAV20L三个数据集的跟踪精度分别提升5.0%、1.3%和4.1%,跟踪成功率分别提升3.8%、2.8%和1.7%,且在对不同跟踪场景实现稳定跟踪的同时能够保持25.5 fps的实时跟踪速度.

概念密度泛函理论及近来的一些进展

概念密度泛函理论(又称密度泛函活性理论或化学密度泛函理论)是密度泛函理论(DFT)的化学活性理论.本文对其理论框架和最新进展进行了概述,包括电负性、硬度、软度、福井函数、亲电性,及其从这些概念中得到的原理.介绍了二元描述符和立体效应定量描述的新进展,并对今后的发展作出了展望.

Studies of the Chemical Reactivity of a Series of Rhodanine Derivatives by Approaches to Quantum Chemistry

This theoretical chemical reactivity study was conducted using the Density Functional Theory (DFT) method, at computational level B3LYP/6-31G (d). It involved a series of six (06) 5-arylidene rhodanines and allowed to predict the chemical reactivity of these compounds. DFT global chemical reactivity descriptors (HOMO and LUMO energies, chemical hardness, softness, electronegativity) were examined to predict the relative stability and reactivity of rhodanin derivatives. Thus, the compound 6 which has an energy gap between the orbitals of ΔEgap = 3.004 eV is the most polarizable, the most reactive, the least stable, the best electron donor and the softest molecule. Calculation of the local indices of reactivity as well as dual descriptors revealed that the sulfur heteroatom of the Rhodanine ring is the privileged site of electrophilic attack in a state of sp3 hybridization and privileged site of nucleophilic attack in a state of sp2 hybridization.

Theoretical Study by Density Functional Theory Method (DFT) of Stability, Tautomerism, Reactivity and Prediction of Acidity of Quinolein-4-One Derivatives

A theoretical study of the reactivity of quinoline-4-one derivatives is undertaken in order to understand the involved mechanisms. The calculations were carried out in gas phase and in N, N-Dimethylformamide (DMF) solution. The Density Functional Theory (DFT) with B3LYP functional associated to 6-311G (d) and 6-311+G (d) bases is used to perform these calculations. The results of the thermodynamic parameters showed that there is an equilibrium relation between the different tautomers. This equilibrium can be used to explain the failure to obtain tetrahydroquinoline from 5,8-dimethoxy-quinolin-4-one. Reactivity analysis from Frontier Molecular Orbitals theory and Fukui function calculations revealed that ketone forms are less reactive than enol ones. The methoxyl substituent decreases the acidity of the nitrogen and oxygen atoms of quinolin-4-one while the bromine increases the acidity of the same sites. These results foresee that nitrogen deprotonation in the case of the brominated compound is easier than in the case of methoxylated ones.

基于SHOT特征提取的三维点云配准算法

点云配准在逆向工程、三维重建和工业检测等领域有着广泛的应用。针对传统的迭代最近点(Iterative Clos⁃est Point,ICP)配准算法中存在对配准位姿要求高且收敛速度慢等问题。论文设计了一种基于方向直方图签名(Unique Sig⁃natures and Histograms for Surface and Texture Description,SHOT)特征提取的三维点云配准算法。利用内部签名描述(Intrin⁃sic Shape Signatures,ISS)算法提取出目标点云的关键点后采用高描述性的SHOT特征描述子对关键点进行特征提取。对关键点进行随机采样后在离线计算好的模板特征中结合二次约束搜索提取出合适的配准点对,利用对偶四元素求解旋转矩阵和平移向量来实现点云的初始配准,最后利用ICP算法实现点云的精确配准。试验结果表明与传统的迭代最近点配准算法、基于快速点特征(Fast Point Feature Histograms,FPFH)配准算法等算法相比,论文算法的配准耗时进一步缩减,对不同位姿下的配准适应性高且对噪声环境下数据缺失有较好的鲁棒性。

预测亲核反应位点方法的比较

预测分子不同位点发生亲核取代反应的活性具有重要的理论和实际意义.目前已提出了许多基于反应物自身电子结构特点的预测方法.本文将碳基化合物、芳香族化合物和吡啶衍生物这3类分子作为测试体系,对14种预测方法的可靠性进行了详细的比较分析.结果表明,体现局部电子软度的方法能很好地预测反应位点,如简缩双描述符方法,但表现静电效应的预测方法,如原子电荷分析和静电势分析,整体表现很差.对于本文中所用的分子体系来讲,简缩双描述符和Hirshfeld电荷分析方法对分子反应位点预测最准确.