人工神经网络在溶解度预测方面的应用

溶解度作为一项重要的物化指标,一直是化学学科的研究重点。然而,通过实验测量获得数据耗时费力,因此,科研人员建立了多种理论方法来进行估算,其中,人工神经网络因其能够关联复杂的多变量情况而受到广泛关注。本文综述了人工神经网络在物质溶解度预测方面的应用,介绍了应用最广泛的3种神经网络(BP神经网络、小波神经网络、径向基神经网络)的模型结构、预测方法和预测优势,探讨了神经网络的不足以及改进方法。文章最后对神经网络在物质溶解度预测方面的发展前景进行了展望。与其他方法相比,人工神经网络技术在物质溶解度预测方面具有预测结果精确度高、操作简单等特点,具有广阔的应用前景,但输入变量选择、隐含层节点数确定、避免局部最优等问题还需逐步建立系统的理论指导。     

The influence of aging on surface free energy of corona treated packaging films

In packaging, plastic films are very often applied as overprinting materials. The printing properties of plastic films depend on the value of the surface free energy. Usually, during storage but before printing, the surface free energy is decreasing as a result of ageing. The aim of this study was to analyse the influence of elevated temperature and UV radiation on ageing properties and variation of the free surface energy for three commercially available plastic films: polyethylene, polypropylene and polyethylene terephthalate. The investigation was done experimentally, and the surface free energy was calculated using two approaches, Owens-Wendt and van Oss-Chaudhury-Good. The time change of polar fractions was also analysed. The calculation results were compared and it was concluded that UV radiation causes more changes in surface free energy than elevated temperature. In some cases, surface free energy values calculated with the applied methods show similar trends.     

人工调控细胞表面受体聚集状态及功能

从利用物理刺激和生物大分子诱导两个方面综述了人工调控细胞表面受体聚集状态的策略. 前者是利用相应的纳米材料在光、 磁场、 温度等物理刺激作用下实现人工调控受体聚集; 后者则利用包括蛋白/多肽类分子、 核酸在内的生物分子的自组装对其靶向识别的受体进行人工调控. 系统介绍了相关研究领域取得的最新进展, 并阐述和展望了该领域现存的挑战和发展方向.     

Effects of thermophoresis particle deposition and of the thermal conductivity in a porous plate with dissipative heat and mass transfer

Network simulation method(NSM) is used to solve the laminar heat and mass transfer of an electricallyconducting,heat generating/absorbing fluid past a perforated horizontal surface in the presence of viscous and Joule heating problem. The governing partial differential equations are non-dimensionalized and transformed into a system of nonlinear ordinary differential similarity equations,in a single independent variable,η. The resulting coupled,nonlinear equations are solved under appropriate transformed boundary conditions. Computations are performed for a wide range of the governing flow parameters,viz Prandtl number,thermophoretic coeffcient(a function of Knudsen number),thermal conductivity parameter,wall transpiration parameter and Schmidt number. The numerical details are discussed with relevant applications. The present problem finds applications in optical fiber fabrication,aerosol filter precipitators,particle deposition on hydronautical blades,semiconductor wafer design,thermo-electronics and problems including nuclear reactor safety.     

Modeling the low-cycle fatigue behavior of visco-hyperelastic elastomeric materials using a new network alteration theory: Application to styrene-butadiene rubber

Although several theories were more or less recently proposed to describe the Mullins effect, i.e. the stress-softening after the first load, the nonlinear equilibrium and non-equilibrium material response as well as the continuous stress-softening during fatigue loading need to be included in the analysis to propose a reliable design of rubber structures. This contribution presents for the first time a network alteration theory, based on physical interpretations of the stress-softening phenomenon, to capture the time-dependent mechanical response of elastomeric materials under fatigue loading, and this until failure. A successful physically based visco-hyperelastic model is revisited by introducing an evolution law for the physical material parameters affected by the network alteration. The general form of the model can be basically represented by two parallel networks: a nonlinear equilibrium response and a time-dependent deviation from equilibrium, in which the network parameters become functions of the damage rate (defined as the ratio of the applied cycle over the applied cycle to failure). The mechanical behavior of styrene-butadiene rubber was experimentally investigated, and the main features of the constitutive response under fatigue loading are highlighted. The experimental results demonstrate that the evolution of the normalized maximum stress only depends on the damage rate endured by the material during the fatigue loading history. The average chain length and the average chain density are then taken as functions of the damage rate in the proposed network alteration theory. The new model is found to adequately capture the important features of the observed stress-strain curves under loading-unloading for a large spectrum of strain and damage levels. The model capabilities to predict variable amplitude tests are critically discussed by comparisons with experiments.     

A novel hybrid immune algorithm and its convergence based on the steepest descent algorithm

This paper proposes a novel hybrid immune algorithm (HIA) that can overcome the typical drawback of the artificial immune algorithm (AIA), which runs slowly and experiences slow convergence. The HIA combines the adaptive AIA based on the steepest descent algorithm. The HIA fully displays global search ability and the global convergence of the immune algorithm. At the same time, it inserts a quasi-descent operator to strengthen its local search ability. A good convergence of the HIA with the quasi-descent idea is shown as well. Numerical experiment results show that the HIA successfully improves running speed and convergence performance.     

Improved delay-dependent stability criterion for neural networks with time-varying delay

In this paper, the problem of delay-dependent asymptotic stability criterion for neural networks with time-varying delay has been considered. A new class of Lyapunov functional which contains a triple-integral term is constructed to derive some new delay-dependent stability criteria. The obtained criteria are less conservative because free-weighting matrices method and a convex optimization approach are considered. Finally, numerical examples are given to illustrate the effectiveness of the proposed method.     

Dynamics about Neural Array with Simple Lateral Inhibitory Connections

Lateral inhibitory effect is a well-known feature of information processing in neural systems. This paper presents a neural array model with simple lateral inhibitory connections. After detailed examining into the dynamics of this kind of neural array, the author gives the sufficient conditions under which the outputs of the network will tend to a special stable pattern called spatial sparse pattern in which if the output of a neuron is 1, then the outputs of the neurons in its neighborhood are 0. This ability called spatial sparse coding plays an important role in self-coding, self-organization and associative memory for patterns and pattern sequences. The main conclusions about the dynamics of this kind of neural array which is related to spatial sparse coding are introduced.