钢筋混凝土梁结构断裂损伤诊断方法

提出一种基于BP神经网络的结构破损诊断方法,该方法以结构破损前后柔度的变化作为破损诊断网络输入,为了解决由于系统响应样本数据空间分布不均匀对网络收敛速度及网络诊断影响问题,对网络训练样本采用广义空间格点进行了交换,模拟算例及应用实例均表明,本文方法能准确诊断结构破损位置与破坏程度,是一种有效的结构破损诊断方法.     

Spatial patterns and environmental risks of ringnet fishing along the Kenyan coast

Ringnet fishing began in the early 20th century and is practised worldwide, mainly to target nearshore pelagic species. The method was introduced to Kenya’s coastal waters by migrant fishers from Tanzania. However, the impacts of this fishing gear remain poorly assessed. We assessed the spatial distribution of ringnet fishing effort and its possible effects on ecosystem components, such as coral reefs, marine megafauna and marine protected areas, on the south coast of Kenya. We tracked 89 ringnet fishing trips made from December 2015 to January 2016 and used spatial multicriteria analysis to determine hotspots of possible environmental risks. The results showed that habitat type and bathymetric profile influenced the spatial distribution of ringnet fishing effort. Mixed seagrass and coral habitats had the highest concentration of the effort. Most of the habitats in the study area were moderately exposed to the impacts of the ringnet fishery. The study identifies high-risk areas that require spatial measures to minimise possible environmental risks of the gear both to habitats and to endangered sea turtles.     

Modeling of Multi-Freedom Ship Motions in Irregular Waves with Fuzzy Neural Networks

1 .IntroductionAshiptravelingatseaundergoesundesirablewave inducedmotions ,namely ,surge ,sway ,heav ing ,rolling ,pitchingandyaw .Thesemotionsoftencauseproblemstothecrew ,theonboardequip mentand ,intheworstcase ,thesafetyofthevessel.Tominimizethewave inducedshipmotions ,controlsystemsmaybeapplied .Theaccuratemodelingofshipmotionsisthereforeveryimportantforshipdesignanddesignofmotioncontrolsystems .Manyresearchershavedevelopedshipmotionpredictionmethodsbasedonthepotentialflowtheo ries (Dong ,…     

Neural Network-Based Active Control for Offshore Platforms

1 .IntroductionWiththedevelopmentofoceantechnology ,moreandmoreextremelylargeandlongflexibleoff shoreplatformsusedforoilexplorationanddrillingoperationarebuiltinhostileoceanenvironments .Ingeneral,thiskindofplatformsisanonlineardistributedparametersystemanditsnaturalfrequencyfallsclosertothedominantwavefrequencieswiththeincreaseofwaterdepth .Besides ,itsstructureisverycomplexandtheexternalwaveforceontheplatformisuncertain .Thus ,theseplatformsarepronetoexcessivewave inducedoscillationsunderbot…     

The Ectoparasitic Gastropod Boonea (= Odostomia) impressa: Population Ecology and the Influence of Parasitism on Oyster Growth Rates

Abstract. Boonea (= Odostomia) impressa is a common ectoparasite of oysters. In the laboratory, small oysters (Crassostrea virginica) parasitized by natural densities of B. impressa produced 75 % less new shell than unparasitized oysters. Shell deposition rates of previously parasitized oysters increased significantly after all B. impressa were removed. Thus, the decrease in growth rate, although significant, apparently was not permanent. B. impressa preferentially parasitized small, living oysters (≤2.5cm) in the field, even though a higher percentage of large, living oysters (>2.5cm) was available. The snails maintained an aggregated distribution on the oyster reef. The number of B. impressa per oyster clump was positively correlated with the number of living oysters per clump, however some clumps with few or no living oysters had many B. impressa. Thus, food availability only partially explained the pattern of distribution. B. impressa was very mobile. About 50 % of the population moved in one week. Reproduction occurred throughout the year with a peak period in May. Recruitment was greatest in July, however new recruits were observed throughout the year. The reduction in growth rate of parasitized oysters, the snaiľs propensity towards parasitizing small oysters and the snail's tendency to be contagiously distributed suggests that B. impressa potentially exerts a significant influence on the population structure and health of oyster populations.     

地形坡度对多金属结核分布的控制作用

选取我国东太平洋多金属结核开辟区内的一个区域作为研究区,利用人工神经网络中应用最为广泛的BP网络,建立控制多金属结核分布的地质因素与多金属结核分布之间的映射关系,探讨地形坡度对多金属结核分布的控制作用.结果表明,在经度、纬度、水深、坡度四个因素中,坡度对结核分布影响程度最大;多金属结核主要分布于坡度小于5°的地方;当坡度小于5°时,丰度与地形坡度呈正相关,品位与地形坡度呈负相关,丰度与品位呈负相关;品位与坡度似呈指数关系,坡度增大,品位降低.     

一种树型网络拓扑简图的生成算法

给出一种树型网络拓扑的语义描述方法及其拓扑简图的生成算法。该方法通过建立网络联接线路标识串集和网络设备连接标识串集，来构造网络设备的拓扑描述，并计算生成网络拓扑简图。     

Kohonen网络与BP网络的集成应用研究

本文介绍了 Kohonen神经网络对输入数据进行聚类方法在卷烟配方中的应用 ,提出了从核心样本动态搜索 BP网络训练样本的新探索 ,摒弃了过去 BP算法中训练样本固定不变 ,互不相交的方法 ,实现了 BP网络和 Kohonen网络动态无缝集成。