基于特征曲线提取的风电数据跨度选取

为减少风力发电技术中海量数据计算容易导致计算灾害的问题,提出一种以有限跨度数据代替海量风电数据的计算方法。通过K-means算法对风电数据进行分类,在分类曲线的基础上对各类风电数据进行特征提取,得到表征特征曲线的最佳类采样天数,最后通过移动"跨度窗"方法求取出能够代替海量风电数据的典型数据跨度。将计算结果应用于某风电场的容量配置计算,通过对比验证了所提出的理论的正确性。

光伏发电数据挖掘中的跨度选取

分析光伏电站有功出力数据时,粒度和跨度是两个重要指标。粒度决定了数据的采样间隔,跨度决定了选取时长。当粒度一定时,跨度过小则包含信息量少,造成信息的缺漏和失真;跨度过大则信息冗余繁杂,造成存储量和计算量不必要的增大。因此,确定数据跨度对光伏出力数据的分析有着重要意义。本文首先对光伏电站出力数据进行统计分析,通过自相关分析得出跨度选取的初步结论。基于天气特征对数据进行聚类,利用统计学原理对不同类别的光伏出力数据进行最优样本容量估计,确定了光伏电站出力数据分析时所需要的数据跨度,并考察了储能容量需求与数据跨度的关系。结果表明,光伏出力数据跨度为31天时即可满足配置储能容量的数据量需求。

小区域混合大跨度数据分类方程的数学稳定性研究

在数学分类问题中,分类方程的稳定性至关重要,当前的数学分类方程对小区域混合大跨度数据集进行分类的过程中,会产生数据混合和波动问题,存在较大的随机性和偏差,无法确保数据分类方程的稳定性。在原有自调节数学分类方程对小区域混合大跨度数据进行分类的基础上,获取该种数据的分类判决方程,采用非线性微分分析方法获取数据分类方程的常数特解,计算方程的积分曲线走向图,用于加强不同小区域混合大跨度数据分类方程解的稳定性。实验结果说明该种方法对小区域混合大跨度数据进行分类的稳定性控制效率和精度都优于传统分类方程,具有较强的数据分类稳定性,取得了令人满意的效果。

Influence of image data set noise on classification with a convolutional network

To evaluate the influence of data set noise, the network in network(NIN) model is introduced and the negative effects of different types and proportions of noise on deep convolutional models are studied. Different types and proportions of data noise are added to two reference data sets, Cifar-10 and Cifar-100. Then, this data containing noise is used to train deep convolutional models and classify the validation data set. The experimental results show that the noise in the data set has obvious adverse effects on deep convolutional network classification models. The adverse effects of random noise are small, but the cross-category noise among categories can significantly reduce the recognition ability of the model. Therefore, a solution is proposed to improve the quality of the data sets that are mixed into a single noise category. The model trained with a data set containing noise is used to evaluate the current training data and reclassify the categories of the anomalies to form a new data set. Repeating the above steps can greatly reduce the noise ratio, so the influence of cross-category noise can be effectively avoided.

Research development of digital maintenance and management system for long-span cable-stayed bridges

Digital cable-stayed bridge maintenance and management system （DCBMS） was developed for the need of maintenance and management of long-span cable-stayed bridges. In this paper, the major functions and theoretical application of eight modules were systematically stated with the background of Harbin Songhua River cable-stayed bridge, which include data management module, inspection and measurement module, assessment module, finite dement analysis module, disease diagnosis and prediction module, maintenance module, query module and help module. By analyzing and calculating the data from manual inspection database, basic database and health monitoring subsystem, DCBMS can accomplish the functions like life prediction, disease diagnosis, comprehensive assessment, maintenance and management of bridges. Therefore, the maintenance and management of long-span cable-stayed bridges can be made digital, professional and scientific. By running this system, a real-time and specific technical guidance can be provided for the maintainers and managers of long- span cable-stayed bridges.

Developing a Coherent Stratigraphic Scheme of the Albertine Graben-East, Africa

The Albertine Graben is one of the most petroliferous onshore rifts in Africa. It forms the northemmost termination of the western arm of the East African Rift System. Its surface exposures were first studied by Wayland [1] and Pickford et al. [2] among others. Pickford et al. [2] especially developed the basic stratigraphic framework of the graben which was later modified by the government geoscientists and international oil companies using subsurface data. However, the stratigraphic units were not fully and formally described, and have been used informally in different and often confusing ways. The current study therefore aims to solve this challenge by establishing a coherent stratigraphic scheme for the entire graben through an integral study of surface and subsurface data. The study involves precise description of the type and reference sections for various formations both in exposure and wells; and has therefore led to the development of lithostratigraphic columns of different basins in the graben. The approach reveals that the Semliki area, south of Lake Albert, has the most complete sedimentary succession in the graben, spanning the period from middle Miocene （ca 15 Ma） to Recent. It also reveals that platform deposits, which form a small fraction of the thickness of the basinal succession, represent a highly condensed sequence which only saw deposition at times of Lake highstand.

上软下硬地层地铁隧道安全覆岩厚度研究

上软下硬土岩复合地层地铁隧道拱顶以上覆岩厚度对围岩自稳起关键控制作用,然而目前关于隧道安全覆岩厚度及其一般规律尚无统一的认识和标准。基于此,文章首先采用有限元强度折减法对不同软弱地层厚度H_(s)和开挖跨度D下的地铁隧道安全覆岩厚度进行研究,得到了地铁隧道临界安全覆岩厚度H_(rcv)与软弱地层厚度Hs间的数学拟合方程,并绘制了临界安全覆岩厚度、软弱地层厚度与开挖跨度“H_(rcv)-H_(s)-D”三维空间分布图;然后通过对青岛地铁隧道施工过程监测数据进行统计分析,验证了隧道围岩自稳性及对周边环境影响的程度与安全覆岩厚度具有良好的相关性。

Analysis and assessment of bridge health monitoring mass data——progress in research/development of “Structural Health Monitoring”

The "Structural Health Monitoring" is a project supported by National Natural Science Foundation for Distinguished Young Scholars of China(Grant No.50725828).To meet the urgent requirements of analysis and assessment of mass monitoring data of bridge environmental actions and structural responses,the monitoring of environmental actions and action effect modeling methods,dynamic performance monitoring and early warning methods,condition assessment and operation maintenance methods of key members are systematically studied in close combination with structural characteristics of long-span cable-stayed bridges and suspension bridges.The paper reports the progress of the project as follows.(1) The environmental action modeling methods of long-span bridges are established based on monitoring data of temperature,sustained wind and typhoon.The action effect modeling methods are further developed in combination with the multi-scale baseline finite element modeling method for long-span bridges.(2) The identification methods of global dynamic characteristics and internal forces of cables and hangers for long-span cable-stayed bridges and suspension bridges are proposed using the vibration monitoring data,on the basis of which the condition monitoring and early warning methods of bridges are developed using the environmental-condition-normalization technique.(3) The analysis methods for fatigue loading effect of welded details of steel box girder,temperature and traffic loading effect of expansion joint are presented based on long-term monitoring data of strain and beam-end displacement,on the basis of which the service performance assessment and remaining life prediction methods are developed.