阵发型神经元网络的联想记忆与分割

以Hindmarsh-Rose(HR)神经元组成网络,研究这些网络的记忆(或模式)存储与时间分割问题.给定一个输入模式,它是几种模式的叠加,网络能够以一部分神经元同步发放的形式一个接一个地分割出每一种模式.如果输入的模式是缺损的,系统能够恢复为原型,即网络具有联想记忆功能.模拟也得到据作者所知至今还未报道的一些现象.     

FitzHugh-Nagumo神经元网络的同步振荡与联想记忆

本文研究由FitzHugh—Nagumo神经元所组成的脉动神经元网络的同步与联想记忆恢复。基于神经元微观生理结构，本文给出具有空间随机分布延时的神经元间耦合，而这种随机分布延时描述了脉动信号从突触前神经元到突触后神经元在轴突上传播所需要的时间。记忆由空时发放的神经元集群表达，在噪声涨落的作用下，系统取得了对不完整输入的记忆恢复。     

双向联想记忆神经网络时滞诱发的同步周期解的计算

双向联想记忆神经网络通常用于描述借助于双向,即前后两个方向,储存或记忆一对类似模式的能力,这方面的研究对于模式识别以及自动控制工程中的应用是至关重要的,而网络系统的同步是近年来国内外研究的热点.本文采用摄动-增量法解析定量地研究了具有四个神经元和两个离散时滞的BAM神经网络模型的同步周期解,不仅提出了所研究的时滞耦合网络系统同步周期解的充分必要条件,而且还给出了由时滞诱发的完全同步周期解的近似解析形式.最后通过数值模拟进一步说明本文所采用的方法是有效的,所得的结果是正确的.本文的结果在设计人工神经网络方面有着潜在的应用.     

基于Hindmarsh-Rose神经元模型的模式分割

联想记忆和模式分割是大脑的重要功能,大脑海马区是这一功能的重要物质基础。Hindmarsh-Rose神经元模型能够较好地模拟大脑海马区神经元的特性。本文以Hindmarsh-Rose神经元为节点,按Hopfield规则构造网络,同时,考虑到大脑中神经元连接的生理学实际情况,通过数值模拟,实现了在弱、中、强三种连接强度情况下的模式分割以及在随机概率连接情况下的模式分割,通过分析比较,提出了进一步改进的方向。     

含时滞反馈与涨落质量的记忆阻尼系统的随机共振

针对具有记忆效应的欠阻尼系统, 存在时滞反馈与涨落质量, 本文主要研究了其输出稳态响应振幅的随机共振效应. 首先通过引入新变量和运用小时滞近似展开理论, 将具有非马尔科夫特性的原系统转化为等价的两维马尔科夫线性系统, 再利用Shapiro-Loginov公式和Laplace变换获得了系统响应的一阶稳态矩和稳态响应振幅的解析表达式. 结果表明: 当系统参数满足Routh-Hurwitz稳定条件时, 稳态响应振幅随质量涨落噪声强度、周期驱动信号频率以及时滞的变化均存在随机共振现象, 其中随机多共振现象也被观察到. 在适当范围内, 通过控制时滞反馈, 系统的随机共振效应随着时滞的增大而增强, 而较长的记忆时间及增大阻尼参数均对共振行为呈现抑制作用.有效调控时滞反馈与记忆效应的变化关系将有助于增强系统对周期驱动信号的响应强度. 最后, 通过数值模拟计算验证了理论结果的有效性.      

Estimation of Attraction Domain and Exponential Convergence Rate of Continuous Feedback Associative Memory

ＩｎｔｒｏｄｕｃｔｉｏｎＤｕｒｉｎｇｔｈｅｐａｓｔｆｅｗｙｅａｒｓ,ｔｈｅａｎａｌｙｓｉｓａｎｄｓｙｎｔｈｅｓｉｓｏｆｃｏｎｔｉｎｕｏｕｓｆｅｅｄｂａｃｋａｓｓｏｃｉａｔｉｖｅｍｅｍｏｒｙｎｅｕｒａｌｎｅｔｗｏｒｋｓｈａｓｂｅｅｎｏｎｅｏｆｔｈｅｍｏｓｔａｃｔｉｖｅａｒｅａｓｏｆｒｅｓｅａｒｃｈａｎｄｈａｓａｔｔｒａｃｔｅｄｔｈｅａｔｔｅｎｔｉｏｎｏｆｍａｎｙｒｅｓｅａｒｃｈｅｒｓ .Ｆｏｒｄｅｔａｉｌｓｏｆｔｈｅｓｅｅｘｉｓｔｉｎｇｒｅｓｕｌｔｓ,ｗｅｃａｎｒｅｆｅｒｔｏ [1- 8]ａｎｄｔｈｅｒｅｆ…     

Effects of internal noise on the spiking regularity of a clustered Hodgkin-Huxley neuronal network

Spiking regularity in a clustered Hodgkin–Huxley (HH) neuronal network has been studied in this letter. A stochastic HH neuronal model with channel blocks has been applied as local neuronal model. Effects of the internal channel noise on the spiking regularity are discussed by changing the membrane patch size. We find that when there is no channel blocks in potassium channels, there exist some intermediate membrane patch sizes at which the spiking regularity could reach to a higher level. Spiking regularity increases with the membrane patch size when sodium channels are not blocked. Namely, depending on different channel blocking states, internal channel noise tuned by membrane patch size could have different influence on the spiking regularity of neuronal networks.     

The race to the nociceptor: mechanical versus temperature effects in thermal pain of dental neurons

The sensing of hot and cold stimuli by dental neurons differs in several fundamental ways. These sensations have been characterized quantitatively through the measured time course of neural discharge signals that result from hot or cold stimuli applied to the teeth of animal models. Although various hypotheses have been proposed to explain the underlying mechanism, the ability to test competing hypotheses against experimental recorded data using biophysical models has been hindered by limitations in our understanding of the specific ion channels involved in nociception of dental neurons. Here we apply recent advances in established biophysical models to test the competing hypotheses. We show that a sharp shooting pain sensa-tion experienced shortly following cold stimulation cannot be attributed to the activation of thermosensitive ion channels, thereby falsifying the so-called neuronal hypothesis,which states that rapidly transduced sensations of coldness are related to thermosensitive ion channels. Our results support a central role of mechanosensitive ion channels and the associated hydrodynamic hypothesis. In addition to the hydrodynamic hypothesis, we also demonstrate that the long time delay of dental neuron responses after hot stimulation could be attributed to the neuronal hypothesis—that a relatively long time is required for the temperature around nociceptors to reach some threshold. The results are useful as a model of how multiphysical phenomena can be combined to provide mechanistic insight into different mechanisms underlying pain sensations.     

双稳系统随机共振数值仿真中信号的过采样

数值仿真方法是非线性双稳系统随机共振研究中经常使用的方法。研究中发现，随机共振数值仿真时需要对信号过采样。但过采样的内在机制是什么，已有研究未能给出解释。通过对双稳系统数值仿真模型的稳定性进行分析，给出了系统的稳定性条件，解释了需要对信号过采样的原因。在此基础，已进行了数值仿真试验，数值仿真结果与理论分析结果相一致，为双稳随机共振系统数值仿真研究时的参数选择提供了依据。     

Qualitative and quantitative evaluation of the interface in activated shape memory alloy composites

The interfacial quality of shape memory alloy (SMA) composites was experimentally evaluated by differential scanning calorimetry, and quantitatively calculated using a combination of Šittner's model and a composite cylinder model. Results have shown that the damage level of the interface in a heating process can be quantitatively evaluated by measuring the transformation enthalpy of the subsequent heating cycle and comparing it with the S-shaped enthalpy kinetic curve. A combination of Šittner's model and a composite cylinder model is capable of calculating both the thermal behavior and the triaxial stress state of SMA composites.     

Three-dimensional modeling and numerical analysis of rate-dependent irrecoverable deformation in shape memory alloys

Shape memory alloys (SMAs) provide an attractive solid-state actuation alternative to engineers in various fields due to their ability to exhibit recoverable deformations while under substantial loads. Many constitutive models describing this repeatable phenomenon have been proposed, where some models also capture the effects of rate-independent irrecoverable deformations (i.e., plasticity) in SMAs. In this work, we consider a topic not addressed to date: the generation and evolution of irrecoverable viscoplastic strains in an SMA material. Such strains appear in metals subjected to sufficiently high temperatures. The need to account for these effects in SMAs arises when considering one of two situations: the exposure of a conventional SMA material (e.g., NiTi) to high temperatures for a non-negligible amount of time, as occurs during shape-setting, or the utilization of new high temperature shape memory alloys (HTSMAs), where the elevated transformation temperatures induce transformation and viscoplastic behaviors simultaneously. A new three-dimensional constitutive model based on established SMA and viscoplastic modeling techniques is derived that accounts for these behaviors. The numerical implementation of the model is described in detail. Several finite element analysis (FEA) examples are provided, demonstrating the utility of the new model and its implementation in assessing the effects of viscoplastic behaviors in shape memory alloys.     

The influence of model parameters and of the thermomechanical coupling on the behavior of shape memory devices

Shape memory materials (SMM) are receiving increasing attention for their use in applications that exploit their dynamic behavior. A thermomechanical model for devices with pseudoelastic behavior has been proposed in previous works [11] (Bernardini and Pence, 2005) [15] (Bernardini and Rega, 2005). The model takes into account several aspects of SMM behavior by means of seven model parameters.In this paper the effect of each parameter on the non-isothermal rate-dependent behavior of the device is studied, by paying particular attention to the effect of the thermomechanical coupling. Some overall synthetic indicators of the behavior of the shape memory device are defined in terms of the model parameters. By evaluating such indicators, a lot of information about the mechanical, thermal and thermomechanical effects on the device behavior can be gained before computing explicitly the response of the shape memory oscillator.The present work may provide a guide for the proper utilization of the model for the investigation of the dynamic response.     

Internal loops in pseudoelastic behaviour of Ti-Ni shape memory alloys: Experiment and modelling

The stress-strain isothermal hysteresis loops due to the incomplete martensitic transformation are analysed for Ti-Ni shape memory alloys. Experiments show the existence of two distinct yield lines for phase transition; one for the forward transformation austenitemartensite (AM), the other for the reverse transformation MA. The tensile behaviour of single crystals with only one yield line (AM) [1] can be considered as an ideal case. An extension of a thermodynamic model for pseudoelasticity [2] allows these two yield lines to be taken into account.

---

Sommario Per leghe Ti-Ni con memoria di forma vengono analizzati i cicli di isteresi isotermici tensione-deformazione prodotti da una incompleta trasformazione martensitica. Gli esperimenti mostrano l'esistenza di due distinte linee di snervamento per la transizione di fase, una verso la trasformazione austenitemartensite (AM), l'altra per la trasformazione inversa MA. Il comportamento a trazione di un singolo cristallo con una sola linea di snervamento (AM) [1], può essere considerato un caso ideale. L'estensione ad un modello termodinamico pseudo-elastico [2] consente di analizzare queste due linee di snervamento.