基于TCMS落地数据的存储、检索、显示软件平台

基于TCMS落地数据的存储、检索、显示软件平台是一套兼顾数据准确性、完整性和实时性的信息传输解决方案,流式处理时序性数据以应用于城市轨道交通行业低延时海量实时数据的特性。该平台由车载数据分类发送、地面服务器解析存储和对外可视化Web服务三部分组成。其中车载数据分类发送作为客户端软件可编译为通用程序包;地面服务端程序处理实时数据流;可视化Web系统主要结合MVC思想(Model_模型,View_视图,Controller_控制器),以Spring Boot开源框架作为基础架构搭建[1]。

风电异常测量数据智能识别方法研究

针对风电测量数据中不同程度存在的异常数据势必影响风电预测精度的问题,结合滑窗方法并运用两种智能算法实现了风电测量数据异常识别。首先分析了风电测量数据的时序性,然后描述了基于高密度连通区域的聚类算法和基于密度的离群点检测算法,并分析了两种方法的性能。最后,针对国内风电场实际测量数据,运用两种算法进行了异常测量数据的识别并在此基础上进行了风电预测,针对计算结果进行了比较分析,验证了所提策略的可行性和有效性。

A Novel Parallel Scheme for Fast Similarity Search in Large Time Series

The similarity search is one of the fundamental components in time series data mining,e.g.clustering,classification,association rules mining.Many methods have been proposed to measure the similarity between time series,including Euclidean distance,Manhattan distance,and dynamic time warping(DTW).In contrast,DTW has been suggested to allow more robust similarity measure and be able to find the optimal alignment in time series.However,due to its quadratic time and space complexity,DTW is not suitable for large time series datasets.Many improving algorithms have been proposed for DTW search in large databases,such as approximate search or exact indexed search.Unlike the previous modified algorithm,this paper presents a novel parallel scheme for fast similarity search based on DTW,which is called MRDTW(MapRedcuebased DTW).The experimental results show that our approach not only retained the original accuracy as DTW,but also greatly improved the efficiency of similarity measure in large time series.

Fortified Financial Forecasting Models Based on Non-Linear Searching Approaches

The paper＇s aim is how to forecast data with variations involving at times series data to get the best forecasting model. When researchers are going to forecast data with variations involving at times series data （i.e., secular trends, cyclical variations, seasonal effects, and stochastic variations）, they believe the best forecasting model is the one which realistically considers the underlying causal factors in a situational relationship and therefore has the best ＂track records＂ in generating data. Paper＇s models can be adjusted for variations in related a time series which processes a great deal of randomness, to improve the accuracy of the financial forecasts. Because of Na＇fve forecasting models are based on an extrapolation of past values for future. These models may be adjusted for seasonal, secular, and cyclical trends in related data. When a data series processes a great deal of randomness, smoothing techniques, such as moving averages and exponential smoothing, may improve the accuracy of the financial forecasts. But neither Na＇fve models nor smoothing techniques are capable of identifying major future changes in the direction of a situational data series. Hereby, nonlinear techniques, like direct and sequential search approaches, overcome those shortcomings can be used. The methodology which we have used is based on inferential analysis. To build the models to identify the major future changes in the direction of a situational data series, a comparative model building is applied. Hereby, the paper suggests using some of the nonlinear techniques, like direct and sequential search approaches, to reduce the technical shortcomings. The final result of the paper is to manipulate, to prepare, and to integrate heuristic non-linear searching methods to serve calculating adjusted factors to produce the best forecast data.

Analysis of seasonal signals and long-term trends in the height time series of IGS sites in China

The seasonal signal and long-term trend in the height time series of 10 IGS sites in China are investigated in this paper. The offset detection and outlier removal as well as the removal of common mode error are performed on the raw GPS time-series data developed by the Scripps Orbit and Permanent Array Center(SOPAC). The seasonal-trend decomposition procedure based on LOESS(STL) is utilized to extract precise seasonal signals, followed by an estimation of the long-term trend with the application of maximum likelihood estimation(MLE) to the seasonally adjusted time series. The Up-compo- nents of all sites are featured by obvious seasonal variations, with significant phase and amplitude modulation on some sites. After Kendall's tau test, a significant trend(99% confidence interval) for all sites is achieved. Furthermore, the trends at sites TCMS and TNML have significant changes at epochs 2009.5384 and 2009.1493(95% confidence interval), respectively, using the Breaks For Additive Seasonal and Trend test. Finally, the velocities and their uncertainties for all sites are estimated using MLE with the white noise plus flicker noise model. And the results are analyzed and compared with those announced by SOPAC. The results obtained in this paper have a higher precision than the SOPAC results.