Descriptions of Protospathidium serpens (Kahl, 1930) and P-fraterculum n. sp (Ciliophora, Haptoria), two species based on different resting cyst morphology

Protospathidium serpens (Kahl, 1930) is frequent in semiterrestrial and terrestrial habitats worldwide. Conventionally, all populations are considered as conspecific because they have very similar overall morphologies and morphometrics. We studied in detail not only the morphology of the vegetative cells but also the resting cysts using live observation, protargol impregnation, and scanning electron microscopy. These revealed a cryptic diversity and biogeographic pattern in details of the dorsal brush and cyst wall morphology. The cyst wall is spiny in the Austrian specimens, while smooth in the South African and Antarctic populations. Accordingly, P. serpens consists of at least two species: P. serpens (with spiny cyst wall) and P. fraterculum n. sp. (with smooth cyst wall); the latter is probably composed of two distinct taxa differing by the absence (South African)/presence (Antarctic) of a monokinetidal bristle tail in brush row 3, the number of dikinetids comprising brush row 1 (seven versus three), and the total number of brush dikinetids (29 versus 17). Protospathidium serpens is neotypitied with the new population from Austria. The significance of resting cyst morphology is discussed with respect to alpha-taxonomy and overall ciliate diversity.

利用BP神经网络实现三维物体姿态的测定

本文利用BP(Back-Propagation)人工神经网络对三维物体的姿态测定进行了研究。姿态测定一直缺少通用而实际的方法,人工神经网络由于具有强大的自组织、自适应学习能力,迅速的并行信息处理能力,可望解决这个问题。但现有BP算法存在训练慢和易陷入局部最小两个问题.本文提出的级联形式网络结构,使BP网络的训练速度大为提高,陷入局部最小的可能性大为降低。利用这种级联结构对飞机模型姿态测定,取得了较好的实验结果。

基于粗糙集-BP神经网络的发动机故障诊断

由于发动机光谱分析监控数据中磨损微粒种类过多,如果将这些微粒信息直接作为神经网络的输入,则存在输入层神经元过多、网络结构复杂等诸多问题。本文将粗糙集引入到发动机故障诊断中来,利用粗糙集在属性约简方面的优势,删除冗余磨损微粒,提取出重要磨损微粒,并将其作为BP神经网络的输入,建立发动机故障诊断模型。该方法降低输入层的神经元个数,简化了网络结构,缩短网络训练时间,并且由于剔除了冗余磨损微粒,减少了由该部分微粒信息不准确而带来的误差,有效提高了故障诊断的精确度。最后通过算例分析验证了相关算法和诊断模型的准确性和有效性。

基于改进的模糊C均值的BP分类器设计

提出了一种基于改进模糊C均值的BP神经网络分类器的设计,通过改进的模糊C均值算法对大量的数据进行聚类划分,然后设计BP神经网络对划分后的数据进行训练和测试,最后由计算机进行综合判断.试验证明该分类器是有效的,可以对高速公路车辆的车型进行迅速判别.

Lewis acid assisted ring-closing metathesis of chiral diallylamines: An efficient approach to enantiopure pyrrolidine derivatives

Lewis acid assisted ring-closing olefin metathesis (RCM) of chiral diallylamines, using the second generation RCM ruthenium-based catalyst, leads to enantiopure pyrrolidine derivatives in 79-93% yields under very mild conditions. The scope of the olefin metathesis has been expanded.