基于个体邻域的改进NSGA-Ⅱ算法

带有精英策略的非支配排序遗传算法（NSGA-II）是在NSGA的基础之上，提出拥挤度和拥挤度比较算子，代替了需要指定共享半径的适应度共享策略，是解决多目标优化问题的经典算法之一。但是NSGA-II算法在保持种群多样性时采取的拥挤距离排挤机制有着pareto前沿分布不均匀的缺陷，因此，提出一种基于个体邻域的改进NSGA-II算法SN-NSGA2。SN-NSGA2将密度聚类算法DBSCAN中邻域的思想应用到排挤机制中去，提出一种个体邻域的构建方法，采用相应的淘汰策略去除个体邻域中的其他邻居个体。实验结果表明相对于NSGA-II算法来说，新算法求出的pareto解集有着更好的分布性以及良好的收敛性。     

改进的排序变异多目标差分进化算法

针对多目标差分进化算法在求解问题时收敛速度慢和均匀性欠佳的问题,提出了一种改进的排序变异多目标差分进化算法（MODE-IRM）。该算法将参与变异的三个父代个体中的最优个体作为基向量,提高了排序变异算子的求解速度;另外,算法采用反向参数控制方法在不同的优化阶段动态调整参数值,进一步提高了算法的收敛速度;最后,引入了改进的拥挤距离计算公式进行排序操作,提高了解的均匀性。采用标准多目标优化问题ZDTl~ZDT4,ZDT6和DTLZ6~DTLZ7进行仿真实验:MODE-IRM在总体性能上均优于MODE-RMO和PlatEMO平台上的MOEA/D-DE、RM-MEDA以及IM-MOEA;在世代距离（GD）、反向世代距离（IGD）和间隔指标（SP）性能度量指标方面,MODE-IRM在所有优化问题上的均值和方差均明显小于MODE-RMO。实验结果表明MODE-IRM在收敛性和均匀性指标上明显优于对比算法。     

Impact of injection limitations on the contact resistance and the carrier mobility of organic field effect transistors

The contact resistance as well as the mobility have developed to key performance indicators for benchmarking organic field-effect transistors. Typically, conventional methods for silicon transistors are employed for their extraction thereby ignoring the peculiarities of organic transistors. This work outlines the required conditions for using conventional extraction techniques for the contact resistance and the mobility based on TCAD simulations and experimental data. Our experimental data contain both staggered and coplanar structures fabricated by exploiting different optimization techniques like SAM treated electrodes, different shearing speeds, PS blending and silicon oxide functionalization. In addition, the work clarifies how injection limited current–voltage characteristics can affect high-performance organic field-effect transistors. Finally, we introduce a semi-physical model for the contact resistance to accurately interpret extracted benchmark parameters by means of the transfer length method (TLM). Guidelines to use conventional extraction techniques with special emphasis on TLM are also provided.     

A complete small-signal HBT model including AC current crowding effect

An improved small-signal equivalent circuit of HBT concerning the AC current crowding effect is proposed in this paper.AC current crowding effect is modeled as a parallel RC circuit composed of Cbi and Rbi,with distributed base-collector junction capacitance also taken into account.The intrinsic portion is taken as a whole and extracted directly from the measured S-parameters in the whole frequency range of operation without any special test structures.An HBT device with a 2 × 20 μm2 emitter-area under three different biases were used to demonstrate the extraction and verify the accuracy of the equivalent circuit.     

基于小生境遗传算法的分类优化方法

对于多分类问题,大多是经二分类器组合进行训练的,在分类类别多、特征维数高时,存在识别准确率不高和训练速度较慢的问题。将超球支持向量机应用到多类问题,为每个类建立一个超球体模型,通过多个超球体划分样本空间。采用改进的基于排挤的小生境遗传算法(improved crowding niche genetic algorithm,ICNGA)进行特征选择,为不同的目标类别寻找最优的特征子集,优化超球支持向量机的输入。利用UCI标准数据集的数值实验表明,在分类数据类别较多、特征维数较高时,经过ICNGA特征选择之后的多超球支持向量机的识别准确度更好,非常适合解决类别数多、特征维数高的分类问题。     

面向反应再生过程的量子粒子群多目标优化

针对催化裂化反应再生过程难以有效解决提升效率、降低损耗、减少排放的多目标优化问题，利用改进的多目标量子粒子群算法进行求解。建立轻油收率、焦炭产率和硫化物排量的多目标优化模型；引入拥挤熵排序更新外部档案，精确估计非支配解集分布性；构造自适应因子以动态调整吸引子，平衡算法的收敛性和多样性；再引入高斯变异进行分段式扰动，增强算法的局部搜索精度，最后求解该优化模型。对某厂催化裂化进行实验，得到轻质油吸收率76.22%，焦炭产率5.72%和硫化物排放量626 mg/m³的结果，均优于其他比较算法，表明改进后的算法可以快速、准确地获得分布均匀的 Pareto 最优解，能有效解决反应再生过程多目标优化问题。     

Design of MCI single and symmetrical on-chip spiral inductors

In this paper, the MCI (multipath crossover interconnection) technique for octagon single and symmetrical spiral inductors has been presented to improve the quality factor. The metal wires of the single and symmetrical inductors formed by the top metal are divided into multiple segments according to the depth of the skin effects. The outermost path of the metal is crossover-interconnected to the innermost path by the underlayer metal and via The crossover technique makes the lengths of the total current paths between two ports approximately equal to each other. Therefore, the induced magnetic flux and resistance of each path can be balanced and the Q-factor of spiral inductors can be enhanced. The proposed MCI technique has been validated by the electromagnetic simulation with the 130-nm 1P6M SiGe BiCMOS process. For the devices with occupying areas of 240×240μm², results of electromagnetic simulation show that about 24% improvement in the Q-peak (3.3 GHz) of the MCI single inductor as compared to conventional single inductors (3.1 GHz), and about 88.1% improvement in the Q-peak (3.2 GHz) of the MCI symmetrical inductor as compared to conventional symmetrical inductors (1.8 GHz).     

Active Site Crowding of Cytochrome P450 3A4 as a Strategy To Alter Its Selectivity

Substrate‐promiscuous enzymes are a promising starting point for the development of versatile biocatalysts. In this study, human cytochrome P450 3A4, known for its ability to metabolise hundreds of drugs, was engineered to alter its regio‐ and stereoselectivity. Rational mutagenesis was used to introduce steric hindrance in a specific manner in the large active site of P450 3A4 and to favour oxidation at a more sterically accessible position on the substrate. Hydroxylation of a synthetic precursor of (R)‐lisofylline, a compound under investigation for its anti‐inflammatory properties, was chosen as a first proof‐of‐principle application of our protein engineering strategy. In a second example, increasing active site crowding led to an incremental shift in the selectivity of oxidation from an internal double bond to a terminal phenyl group in a derivative of theobromine. The same correlation between crowding and selectivity was found in a final case focused on the hydroxylation of the steroid sex hormone progesterone.     

Potential applications of luminescent molecular rotors in food science and engineering

Fluorescent molecular rotors (MRs) are compounds whose emission is modulated by segmental mobility; photoexcitation generates a locally excited (LE), planar state that can relax either by radiative decay (emission of a photon) or by formation of a twisted intramolecular charge transfer (TICT) state that can relax nonradiatively due to internal rotation. If the local environment around the probe allows for rapid internal rotation in the excited state, fast non-radiative decay can either effectively quench the fluorescence or generate a second, red-shifted emission band. Conversely, any environmental restriction to twisting in the excited state due to free volume, crowding or viscosity, slows rotational relaxation and promotes fluorescence emission from the LE state. The environmental sensitivity of MRs has been exploited extensively in biological applications to sense microviscosity in biofluids, the stability and physical state of biomembranes, and conformational changes in macromolecules. The application of MRs in food research, however, has been only marginally explored. In this review, we summarize the main characteristics of fluorescent MRs, their current applications in biological research and their current and potential applications as sensors of physical properties in food science and engineering.     

基于余弦距离的多目标粒子群优化算法

针对粒子群优化算法具有的个体分布不均匀以及重复个体较多等缺陷,提出了一种基于余弦距离的多目标粒子群优化算法,该算法根据外部精英存储策略,利用余弦距离排挤机制来选取最分散的粒子,扩大 Pareto最优解集的收敛性和多样性,增强算法的全局寻优能力。通过采用标准多目标优化问题ZDTl～ZDT3进行仿真实验与粒子群算法、混沌粒子群算法、基于拥挤距离的多目标优化算法对比表明,该算法在Pareto前沿的收敛性和多样性方面均优于基于拥挤距离排挤机制,并具有较高的效率