支持向量机在边坡稳定性预测中的应用

针对离子型稀土原地浸矿边坡稳定性问题的小样本、非线性等特点,采用一种边坡稳定性预测模型——支持向量机(SVM)模型,分别使用网格参数寻优、遗传算法(GA)参数寻优、粒子群(PSO)寻优算法来计算SVM模型的参数,分析比较3种算法得到的参数,确定最适合离子型稀土原地浸矿边坡稳定性预测的SVM回归模型的参数寻优算法。     

Employees' self‐efficacy and perception of individual learning in teams: The cross‐level moderating role of team‐learning behavior

Despite the importance of employee learning for organizational effectiveness, scholars have yet to identify the factors that influence employees' perception of individual learning. This paper identified employees' self‐efficacy as a potential antecedent to their perception of individual learning in the context of teamwork. We also hypothesized that team‐learning behavior had a moderating effect on the relationship between employees' self‐efficacy and their perception of individual learning. We conducted a study of 236 teams working in a retail firm, comprising 236 team supervisors and 1397 employees, and analyzed the data using hierarchical linear modeling. This study revealed that employees' individual‐level self‐efficacy was positively associated with their perception of individual learning in teams. Additionally, team‐learning behaviors moderated the positive relationship between employees' self‐efficacy and the perception of individual learning. This study has theoretical and practical implications for a more nuanced understanding of the perception of individual learning in the context of teamwork. Copyright © 2016 John Wiley & Sons, Ltd.     

基于相关向量机模型的腐蚀声发射信号识别

相关向量机(RVM)模型的分类性能与其核函数参数的选择有密切关系。本文分别利用人工蜂群算法(ABC)、粒子群算法(PSO)和遗传算法(GA)寻找相关向量机模型的最优参数,对几种方法的寻优性能进行了对比。采用基于二叉树结构的一对多扩展方法,对二分类相关向量机模型进行了扩展,建立了四分类模型。基于该分类模型对罐底腐蚀声发射信号进行识别,将声发射特征参数和频域参数作为模型的输入参数,获得了较好的识别结果。     

Can Desalination and Clean Energy Combined Help to Alleviate Global Water Scarcity?

The major present hindrance in using desalination to help alleviate global water scarcity is the cost of this technology, which, in turn is due to energy cost involved. This study examines historical trends in desalination and breaks up the cost of desalination into energy based and nonenergy based. It then develops the learning curves (relationship between cumulative production and market price) for desalination. Assuming that the photovoltaic (PV) technology will be the dominant form of energy used in the desalination process, the existing PV learning curve and desalination learning curve are combined to explore the viability of large‐scale adoption of desalination in the future. The world has been divided into seven regions and it is assumed that water demand from desalinated water will be met only within the 100‐km coastal belt. It is shown that, in most of the regions, other than sub‐Saharan Africa, Central America, and South Asia (where water tariffs are low), the desalination (without considering energy) becomes viable by 2040. For PV technology, less than 1 million MW per annum growth is required till 2050 to make it affordable. Globally, desalination with renewable energy can become a viable option to replace domestic and industrial water demand in the 100‐km coastal belt by 2050.     

基于深度学习方法的PM2.5精细化时空估算模型

提出一种基于深度学习方法的地面PM_2.5浓度时空估算模型（PM_2.5-DNN）,该模型基于葵花-8卫星反演的AOD数据,结合PM_2.5监测站和气象站点观测数据对北京市地面PM_2.5浓度进行了逐时的高精度模拟,同时将PM_2.5-DNN模型的模拟性能与当前的主流方法进行了对比研究.结果表明,使用PM_2.5-DNN模型估算的北京地区1km分辨率每小时地面PM_2.5浓度与地表监测站观测数据对比的一致性较好,模型估算精度可达到R²=0.88,性能优于当前的主流方法.本文所提出的方法适用于区域尺度PM_2.5浓度时空分布细粒度建模与估算,采用端到端的训练方式构建模型,为精细的PM_2.5浓度估算提供了一个简便而有效的方法模型.     

Computer-based training for safety: comparing methods with older and younger workers

INTRODUCTION: Computer-based safety training is becoming more common and is being delivered to an increasingly aging workforce. Aging results in a number of changes that make it more difficult to learn from certain types of computer-based training. Instructional designs derived from cognitive learning theories may overcome some of these difficulties. METHODS: Three versions of computer-based respiratory safety training were shown to older and younger workers who then took a high and a low level learning test. RESULTS: Younger workers did better overall. Both older and younger workers did best with the version containing text with pictures and audio narration. DISCUSSION: Computer-based training with pictures and audio narration may be beneficial for workers over 45 years of age. IMPACT ON INDUSTRY: Computer-based safety training has advantages but workers of different ages may benefit differently. Computer-based safety programs should be designed and selected based on their ability to effectively train older as well as younger learners.     

模拟水下近距非接触爆炸环境的中量级冲击试验

由于非接触爆炸冲击环境引起的冲击值的范围取决于单位时间内的位移,中量级冲击试验条件是建立在舰船、装备受到剧烈的冲击载荷而不至于毁灭的基础上的,因此,采用了中量级冲击机进行试验,介绍了试验条件及参数。试验结果表明,用中量级冲击机来模拟水下近距离非接触爆炸性环境并对装备进行考核是当前可行的一种技术手段。     

VOCs、细菌及颗粒物混合暴露致小鼠学习记忆障碍机制研究

室内空气污染的现状是多种污染源共存,挥发性有机物(volatile organic compounds, VOCs)、细菌及颗粒物等是典型室内空气污染物。神经行为学毒性是室内空气污染引起的敏感毒性作用之一。为了探讨VOCs、细菌及颗粒物混合暴露对小鼠学习记忆能力的影响及机制,选用雄性昆明小鼠70只,随机分为对照(G1)和2～7(G2～G7)号染毒组。采用水迷宫和抓力仪测定小鼠的学习记忆潜伏期和抓力,染毒结束后测定全脑中活性氧(ROS)、丙二醛(MDA),神经递质谷氨酸(Glu)、乙酰胆碱(Ach)含量以及胆碱能系统的乙酰胆碱转移酶(ChAT)和乙酰胆碱酯酶(TChE)活力,同时分析脑源性神经营养因子(BDNF)、胶质细胞神经营养因子(GDNF)及神经生长因子(NGF)的水平。结果显示,第6天,G5、G6及G7小鼠的抓力、逃避潜伏期及在原平台所在象限的探索时间较对照组存在显著差异,并且伴有ROS、MDA含量的显著升高,Glu含量的显著升高,Ach含量、ChAT及TChE活力的显著降低,以及神经营养因子的显著下调(P<0.05或P<0.01)。研究结果表明,VOCs、颗粒物及细菌混合暴露能够导致小鼠学习记忆障碍,混合暴露引起的氧化损伤诱导的神经兴奋或抑制性毒性作用,以及神经营养因子沿轴突逆向传递降低或中断两方面作用导致神经递质产生和释放异常,进而引起学习记忆能力降低。     

中国高新技术产业能源消耗与碳排放分析

基于2000-2007年中国高新技术产业经济发展与能源消耗(简称能耗)数据,统计分析了中国高新技术产业能耗强度变化趋势,并采用“学习曲线”方法论证了高新技术产业能耗的特征;在此基础上,采用政府间气候变化专门委员会(IPCC)温室气体指导方针的方法核算了中国高新技术产业的碳排放量及排放强度.结果表明:(1)中国高新技术产...