City living and urban upbringing affect neural social stress processing in humans

More than half of the world's population now lives in cities, making the creation of a healthy urban environment a major policy priority. Cities have both health risks and benefits, but mental health is negatively affected: mood and anxiety disorders are more prevalent in city dwellers and the incidence of schizophrenia is strongly increased in people born and raised in cities. Although these findings have been widely attributed to the urban social environment, the neural processes that could mediate such associations are unknown. Here we show, using functional magnetic resonance imaging in three independent experiments, that urban upbringing and city living have dissociable impacts on social evaluative stress processing in humans. Current city living was associated with increased amygdala activity, whereas urban upbringing affected the perigenual anterior cingulate cortex, a key region for regulation of amygdala activity, negative affect and stress. These findings were regionally and behaviourally specific, as no other brain structures were affected and no urbanicity effect was seen during control experiments invoking cognitive processing without stress. Our results identify distinct neural mechanisms for an established environmental risk factor, link the urban environment for the first time to social stress processing, suggest that brain regions differ in vulnerability to this risk factor across the lifespan, and indicate that experimental interrogation of epidemiological associations is a promising strategy in social neuroscience.     

Large recurrent microdeletions associated with schizophrenia

Reduced fecundity, associated with severe mental disorders, places negative selection pressure on risk alleles and may explain, in part, why common variants have not been found that confer risk of disorders such as autism, schizophrenia and mental retardation. Thus, rare variants may account for a larger fraction of the overall genetic risk than previously assumed. In contrast to rare single nucleotide mutations, rare copy number variations (CNVs) can be detected using genome-wide single nucleotide polymorphism arrays. This has led to the identification of CNVs associated with mental retardation and autism. In a genome-wide search for CNVs associating with schizophrenia, we used a population-based sample to identify de novo CNVs by analysing 9,878 transmissions from parents to offspring. The 66 de novo CNVs identified were tested for association in a sample of 1,433 schizophrenia cases and 33,250 controls. Three deletions at 1q21.1, 15q11.2 and 15q13.3 showing nominal association with schizophrenia in the first sample (phase I) were followed up in a second sample of 3,285 cases and 7,951 controls (phase II). All three deletions significantly associate with schizophrenia and related psychoses in the combined sample. The identification of these rare, recurrent risk variants, having occurred independently in multiple founders and being subject to negative selection, is important in itself. CNV analysis may also point the way to the identification of additional and more prevalent risk variants in genes and pathways involved in schizophrenia.     

钢包精炼炉的电极系统智能建模及控制

针对钢包精炼炉电极控制系统具有非线性、时变、模型不确定、大滞后、多输入多输出耦合的特点,提出一种基于神经网络实时在线辩识的内模控制方案.控制器采用神经网络解耦,将混沌机制引入到BP算法中,用以加快学习的收敛速度.仿真结果证实了控制策略的有效性.     

一个基于人工神经网络的新一代精神分裂症诊断系统ESSD

本文介绍一个利用人工神经网络进行分型和自动知识获取的新一代精神分裂症诊断系统。阐述了训练样本的统计方法，神经网络的模糊与清晰性相结合的分型方法以及动态自学习方法，在输出规则中引入了黑洞束缚的概念。     

基于人工神经元网络模型的预测控制研究

研究了基于人工神经元网络模型的非线性预测控制，所采用的网络为一种将线性模型与多层前向网络相结合的ＤＬＦ网络，仿真结果表明，该“混合网络”易训练，收敛速度可大大加快，在ＤＬＦ模型的基础上，本文研究了一种非线性预测控制算法，它的显著特点是在线计算量小。对于一非线性过程－球形罐液位的仿真结果表明，基于ＤＬＦ的非线性预测控制效果颇佳。     

Neurogenic hypothesis and psychiatric disorders

Psychiatric illness, such as affective disorders, anxiety disorders and schizophrenia, exerts exceptional personal burden on affected individuals. Although not physically noticeable, these disorders cost enormously on ones’ family and society. Currently pharma-ceutical and psychological treatments are generally accepted as effective for psychiatric disorders, while the exact mechanisms underlying the treatment efficacy, etiology and neurobiology of the disorders remain elusive. In the past decade, "neurogenic hy-pothesis" emerged as an attempt to explain the nature of psychiatric illness. The origination of the hypothesis is based on several pre-clinical and clinical observations. First, stress, which is a common risk factor of the disorders, was found to suppress neurogenesis; second, treatment for the illnesses like antidepressants and antipsychotics were shown to improve neurogenesis and behavioral deficits simultaneously; and third, the therapeutic effect of antidepressants was abolished in animal models when neuro-genesis was blocked. Increasing efforts were invested to determine whether neurogenesis is a key to the understanding and treatment of psychiatric disorders, although contrasting results are also found and thus the importance of neurogenesis remains a matter of debate. The present chapter will discuss the recent findings about the involvement of neurogenesis in major depression, anxiety disorders and schizophrenia, and whether neurogenesis would be a potential target for development of the treatment in the future.     

利用串/并联神经网络建立化学反应器模型

本文研究使用线性动态神经网络与非线性的静态网络相结合的混合建模方式解决复杂非线性系统的建模问题。使用混合神经网络建模，可以降低单个网络的训练难度，基于此，也可将非线性系统控制策略的求解分解，转换为线性系统的求解。从而改善使用单一神经网络建模存在的精度不高以及训练时间长等不足，也为非线性系统控制策略的求解提供方便。本文以一个典型多变量系统——连续搅拌釜式反应器（CSTR）作为仿真对象，详细研究和实现了两类神经网络串联和并联的混合建模方法，并对结果进行了比较。     

立窑水泥煅烧过程的控制策略

从机械化立窑水泥煅烧过程的特点及其控制系统的现状出发,运用神经网络和模糊控制技术,提出基于神经网络的立窑水泥煅烧过程的模糊控制策略。据此,设计出以控制策略为核心的立窑水泥煅烧过程自组织控制器,系统仿真及运行结果表明,该控制系统具有很好的鲁棒性、稳定性,其控制效果较为满意。     

基于AEPSO优化的神经网络多步预测控制

为提高神经网络预测控制的性能,提出了基于自适应扩展粒子群优化的神经网络预测控制方案。基本PSO算法中,每个粒子的更新受粒子个体极值和局部极值的影响,为了提高其全局收敛性,采用多粒子策略,使每个粒子的更新受更多其他粒子的影响;为提高收敛速度,采用自适应策略,对参数c0进行自适应调整,使c0随着迭代次数的增加而逐渐减小,这样,在PSO算法的搜索过程中,随着迭代次数的增加,搜索区域会越来越小,从而加快PSO算法收敛速度。运用该算法实现神经网络预测控制中的滚动优化,在有限时域内对控制序列进行寻列,给出基于粒子群优化的神经网络预测控制系统的稳定性证明。仿真结果表明,基于粒子群优化的神经网络预测控制系统具有良好的跟踪性能。     

利用RBF神经网络实现聚合反应的内模控制

文中研究基于径向基(RBF)神经网络算法的内模控制策略在苯乙烯本体聚合反应相对分子质量分布控制领域的应用。利用神经网络对非线性系统的逼近能力，把内模控制推广到聚合反应过程质量指标控制这一非线性系统中。针对建模过程中存在的稳态误差，在训练数据中增加了部分静态数据，有效的提高了模型的验证精度，大大改善了由神经网络构成的内模控制器的控制精度，消除了系统余差。仿真结果证明，基于神经网络算法的内模控制策略达到了较好的控制质量。     

高斯激励混沌神经元系统及其应用

突破以往混沌神经元模型以Sigmoid函数作为激励函数的过程,构建了由高斯函数独自作为激励函数的混沌神经元模型,分析了它的混沌动力学特性;撤销模拟退火策略后,通过对时间序列的重要指标,如功率谱及最大Lyapunov指数的分析,证实高斯激励的神经元动力系统能够保持永久的混沌搜索状态;利用该系统对灰度图像进行加密,阐述了其原理及算法,通过对加密前后直方图的考查,说明了该混沌加密算法具有较强的抗统计分析能力.     

基于环境复杂度的危险品运输车辆碰撞预警策略

为研究传统车辆碰撞预警系统误警率较高、不完全适用于危险品运输车的问题，文中在碰撞预警策略中引入环境复杂度的概念。对静态复杂度和动态复杂度影响因子进行具体分析，提出了危险判别指标的计算方法和预警策略的具体框架，在此基础上结合国内公开的运输载货车事故报告及真实运输路段的环境信息，利用神经网络对非线性关系的高拟合度建立环境复杂度量化模型，收集真实路段下带有普通预警系统的危险品运输车辆实车数据并进行策略验证。研究结果表明:相较于市面上普通的碰撞预警系统，所提出的预警策略误警率从41％降低至8％，分类后的正确率达92％，可以较好地适应驾驶人对危险的判别和制动习惯，提高危险品运输车辆运行的安全性。      

一种混沌神经网络及在优化计算中的应用

提出了一种混沌神经网络模型。在Hopfield网络中引入混沌机制,首先在混沌动态下搜索,然后利用HNN梯度优化搜索。对非线性函数的优化问题仿真表明算法具有很强的克服陷入局部极小能力。     

中国海洋碳汇交易市场构建

 分析了发展海洋碳汇市场交易的优劣势,从海洋碳汇交易市场的构建原则、模式选择、发展路径、要素设计、运行机制及保障措施等方面,综合分析探讨了中国海洋碳汇交易市场的构建策略。结果表明：（1）中国海洋碳汇交易市场的构建应采取“市场+政府”的模式,同时市场建设应采取分阶段发展的路径安排;（2）海洋碳汇交易市场建设内容包括市场要素和供求、价格、风险等市场运行机制及支撑机制;（3）为保障海洋碳汇交易市场能够持续、有效地运行,需加强组织领导、完善体制机制,加强顶层设计、完善法律体系,深化科学研究、提高技术水平,注重人才培养、强化智力支持。     

连续发酵过程的神经网络预测和控制

给出了神经网络趋化性算法的一种新的实现策略，在此基础上，提出了一种动态递归神经网络建模方法和一种控制作用受限的自学习非线性控制方法。将其用于连续搅拌签式发酵器的状态变量的在线预测和优化控制，仿真结果表明，预测精度高，控制效果好，具有强抗扰和强鲁棒性。在不知道生化过程模型结构的情况下，神经网络模型，可取很容易地通过在线或离线学习到高度复杂的非线性生化过程的输入．输出关系。对于经过最优操作点，稳态增益的符号会发生变化的这类难以控制的生化过程，神经网络非线性控制策略，可以使生化反应器始终维持在最优状况。本方法有望在实际工业过程中得到应用。     

Lesions of the human amygdala impair enhanced perception of emotionally salient events

Commensurate with the importance of rapidly and efficiently evaluating motivationally significant stimuli, humans are probably endowed with distinct faculties and maintain specialized neural structures to enhance their detection. Here we consider that a critical function of the human amygdala is to enhance the perception of stimuli that have emotional significance. Under conditions of limited attention for normal perceptual awareness-that is, the attentional blink-we show that healthy observers demonstrate robust benefits for the perception of verbal stimuli of aversive content compared with stimuli of neutral content. In contrast, a patient with bilateral amygdala damage has no enhanced perception for such aversive stimulus events. Examination of patients with either left or right amygdala resections shows that the enhanced perception of aversive words depends specifically on the left amygdala. All patients comprehend normally the affective meaning of the stimulus events, despite the lack of evidence for enhanced perceptual encoding of these events in patients with left amygdala lesions. Our results reveal a neural substrate for affective influences on perception, indicating that similar neural mechanisms may underlie the affective modulation of both recollective and perceptual experience.     

某钢厂冷轧发货预报模型的预测算法及策略

分析了简单趋势、时间序列、神经网络等预测方法，结合组合预测和滚动优化的思想提出了适合于某钢厂冷轧发货预报的预测策略，给出了严格的基于实际生产数据的仿真研究。仿真结果表明，依据模型提出的预测策略是可行的。     

非高斯随机分布系统的集成故障诊断与容错控制

针对一类非高斯非线性随机分布系统,提出了一种集成故障诊断与容错控制算法. 将基于有理平方根模型逼近其系统输出概率密度函数（PDF）,在此基础上给出了基于RBF神经网络观测器的故障诊断算法,诊断出系统发生的渐变故障信息,基于Lyapunov稳定性定理对其观测误差系统进行收敛性分析. 根据故障诊断信息,给出了PI跟踪容错控制策略,使得系统输出概率密度函数仍能够跟踪给定的分布. 仿真结果验证了该集成故障诊断和容错控制算法的有效性.      

基于改进卷积神经网络与动态衰减学习率的环境声音识别算法

环境声音识别在音频检索、监控方面有着广泛的应用,是听觉识别任务中的一个热门研究领域。但由于其声音信号的复杂多变,使得该任务在识别率提升方面依然面临许多挑战。针对这一问题本文提出了基于改进卷积神经网络的环境识别模型（S-CNN）,该模型采用反复堆叠的递减型卷积核提取不同尺度的局部特征,并在每层卷积层后采用Batch Normalization（BN）层对特征进行归一化操作。同时,利用动态衰减的学习率训练模型,以提高模型收敛速度与收敛稳定性。实验结果表明,相比于传统的机器学习与卷积神经网络模型,本文所设计的改进卷积神经网络模型S-CNN具有更好的识别率。在ESC-10环境声音数据库上,识别精度达到91.3%。