Probabilistic Automatic Outlier Detection for Surface Air Quality Measurements from the China National Environmental Monitoring Network

Although quality assurance and quality control procedures are routinely applied in most air quality networks, outliers can still occur due to instrument malfunctions, the influence of harsh environments and the limitation of measuring methods. Such outliers pose challenges for data-powered applications such as data assimilation, statistical analysis of pollution characteristics and ensemble forecasting. Here, a fully automatic outlier detection method was developed based on the probability of residuals, which are the discrepancies between the observed and the estimated concentration values. The estimation can be conducted using filtering—or regressions when appropriate—to discriminate four types of outliers characterized by temporal and spatial inconsistency, instrument-induced low variances, periodic calibration exceptions, and less PM_(10) than PM_(2.5) in concentration observations, respectively. This probabilistic method was applied to detect all four types of outliers in hourly surface measurements of six pollutants(PM_(2.5), PM_(10),SO_2,NO_2,CO and O_3) from 1436 stations of the China National Environmental Monitoring Network during 2014-16. Among the measurements, 0.65%-5.68% are marked as outliers. with PM_(10) and CO more prone to outliers. Our method successfully identifies a trend of decreasing outliers from 2014 to 2016,which corresponds to known improvements in the quality assurance and quality control procedures of the China National Environmental Monitoring Network. The outliers can have a significant impact on the annual mean concentrations of PM_(2.5),with differences exceeding 10 μg m^(-3) at 66 sites.

Exploring spatial non-stationarity of near-miss ship collisions from AIS data under the influence of sea fog using geographically weighted regression:A case study in the Bohai Sea,China

Sea fog is a disastrous weather phenomenon,posing a risk to the safety of maritime transportation.Dense sea fogs reduce visibility at sea and have frequently caused ship collisions.This study used a geographically weighted regression(GWR)model to explore the spatial non-stationarity of near-miss collision risk,as detected by a vessel conflict ranking operator(VCRO)model from automatic identification system(AIS)data under the influence of sea fog in the Bohai Sea.Sea fog was identified by a machine learning method that was derived from Himawari-8 satellite data.The spatial distributions of near-miss collision risk,sea fog,and the parameters of GWR were mapped.The results showed that sea fog and near-miss collision risk have specific spatial distribution patterns in the Bohai Sea,in which near-miss collision risk in the fog season is significantly higher than that outside the fog season,especially in the northeast(the sea area near Yingkou Port and Bayuquan Port)and the southeast(the sea area near Yantai Port).GWR outputs further indicated a significant correlation between near-miss collision risk and sea fog in fog season,with higher R-squared(0.890 in fog season,2018),than outside the fog season(0.723 in non-fog season,2018).GWR results revealed spatial non-stationarity in the relationships between-near miss collision risk and sea fog and that the significance of these relationships varied locally.Dividing the specific navigation area made it possible to verify that sea fog has a positive impact on near-miss collision risk.

Automatic Abnormal Electroencephalograms Detection of Preterm Infants

Many preterm infants suffer from neural disorders caused by early birth complications. The detection of children with neurological risk is an important challenge. The electroencephalogram is an important technique for establishing long-term neurological prognosis. Within this scope, the goal of this study is to propose an automatic detection of abnormal preterm babies’ electroencephalograms (EEG). A corpus of 316 neonatal EEG recordings of 100 infants born after less than 35 weeks of gestation were preprocessed and a time series of standard deviation was computed. This time series was thresholded to detect Inter Burst Intervals (IBI). Temporal features were extracted from bursts and IBI. Feature selection was carried out with classification in one step so as to select the best combination of features in terms of classification performance. Two classifiers were tested: Multiple Linear Regressions and Support Vector Machines (SVM). Performance was computed using cross validations. Methods were validated on a corpus of 100 infants with no serious brain damage. The Multiple Linear Regression method shows the best results with a sensitivity of 86.11% ± 10.01%, a specificity of 77.44% ± 7.62% and an AUC (Area under the ROC curves) of 0.82 ± 0.04. An accurate detection of abnormal EEG for preterm infants is feasible. This study is a first step towards an automatic analysis of the premature brain, making it possible to lighten the physician’s workload in the future.

Aware Time in Bug Fix—A Novel Automatic Test Case Selection for Prioritization of Version Control

The supreme goal of the Automatic Test case selection techniques is to guarantee systematic coverage, to recognize the usual error forms and to lessen the test of redundancy. It is unfeasible to carry out all the test cases consistently. For this reason, the test cases are picked and prioritize it. The major goal of test case prioritization is to prioritize the test case sequence and finds faults as early as possible to improve the efficiency. Regression testing is used to ensure the validity and the enhancement part of the changed software. In this paper, we propose a new path compression technique (PCUA) for both old version and new version of BPEL dataset. In order to analyze the enhancement part of an application and to find an error in an enhancement part of an application, center of the tree has been calculated. Moreover in the comparative analysis, our proposed PCUA- COT technique is compared with the existing XPFG technique in terms of time consuming and error detection in the path of an enhancement part of BPEL dataset. The experimental results have been shown that our proposed work is better than the existing technique in terms of time consuming and error detection.

Automated regression test method for scientific computing libraries:Illustration with SPHinXsys

Scientific computing libraries,whether in-house or open-source,have witnessed enormous progress in both engineering and scientific research.Therefore,it is important to ensure that modifications to the source code,prompted by bug fixing or new feature development,do not compromise the accuracy and functionality that have been already validated and verified.This paper introduces a method for establishing and implementing an automatic regression test environment,using the open-source multi-physics library SPHinXsys as an illustrative example.Initially,a reference database for each benchmark test is generated from observed data across multiple executions.This comprehensive database encapsulates the maximum variation range of metrics for different strategies,including the time-averaged,ensemble-averaged,and dynamic time warping methods.It accounts for uncertainties arising from parallel computing,particle relaxation,physical instabilities,and more.Subsequently,new results obtained after source code modifications undergo testing based on a curve-similarity comparison against the reference database.Whenever the source code is updated,the regression test is automatically executed for all test cases,providing a comprehensive assessment of the validity of the current results.This regression test environment has been successfully implemented in all dynamic test cases within SPHinXsys,including fluid dynamics,solid mechanics,fluid-structure interaction,thermal and mass diffusion,reaction-diffusion,and their multi-physics couplings,and demonstrates robust capabilities in testing different problems.It is noted that while the current test environment is built and implemented for a particular scientific computing library,its underlying principles are generic and can be easily adapted for use with other libraries,achieving equal effectiveness.

基于GPRS无线通讯技术的自动化灌溉系统设计

针对传统的大水漫灌等灌溉方式灌水不均、容易造成农作物病害或涝死、浪费水资源和人工成本较高的问题,基于GPRS无线通讯技术对自动化灌溉系统进行了设计。为了获取有效的灌溉数据,同时能够对数据进行统计、分析和预测,设计了自动灌溉数据信息的预处理方法,并采用多元线性回归预测模型对灌溉数据进行预测。为了验证该自动化灌溉系统的性能,对其进行了数据采集试验和灌溉预测试验,结果表明:系统对灌溉数据监测和预测的准确率均较高。

基于神经核网络高斯过程回归的甲板运动预测

甲板运动预测与补偿是舰载机自动着舰的关键技术之一。传统甲板运动预测方法依赖于运动建模的准确性和参数调整,面临复杂海况、不同舰型、航态变化时具有适应性差、预测时长短、结果可靠性低等问题。提出一种基于神经核网络高斯过程回归(NKN-GPR)的甲板运动预测模型,使用神经核网络(NKN)实现高斯过程回归(GPR)模型自动复合核构造,有效改善基于规则库自动核搜索(ACKS)算法依赖人工先验知识的不足。以正弦波组合模型和功率谱模型构造仿真数据,对NKN-GPR模型和基于最小二乘法的自回归(AR)模型进行对比仿真验证,仿真结果表明,NKN-GPR模型在运动预测精度、平滑性、预测时长等方面具有显著优势,证明了所提算法的有效性,可为舰载机自动安全着舰提供理论支撑。

Deep Structure Optimization for Incremental Hierarchical Fuzzy Systems Using Improved Differential Evolution Algorithm

The optimization of the rule base of a fuzzy logic system (FLS) based on evolutionary algorithm has achievednotable results. However, due to the diversity of the deep structure in the hierarchical fuzzy system (HFS) and thecorrelation of each sub fuzzy system, the uncertainty of the HFS’s deep structure increases. For the HFS, a largenumber of studies mainly use fixed structures, which cannot be selected automatically. To solve this problem, thispaper proposes a novel approach for constructing the incremental HFS. During system design, the deep structureand the rule base of the HFS are encoded separately. Subsequently, the deep structure is adaptively mutated basedon the fitness value, so as to realize the diversity of deep structures while ensuring reasonable competition amongthe structures. Finally, the differential evolution (DE) is used to optimize the deep structure of HFS and theparameters of antecedent and consequent simultaneously. The simulation results confirm the effectiveness of themodel. Specifically, the root mean square errors in the Laser dataset and Friedman dataset are 0.0395 and 0.0725,respectively with rule counts of rules is 8 and 12, respectively.When compared to alternative methods, the resultsindicate that the proposed method offers improvements in accuracy and rule counts.

基于遗传算法-v支持向量回归的船舶轨迹预测

为了提高船舶轨迹预测精度,避免船舶海上航行事故的发生,本文采用遗传算法对v-支持向量回归进行参数寻优,以此来分别构建关于经纬度的船舶轨迹预测模型。选取水上移动业务标识码为356772000的货船在2022年6月的船舶自动识别系统数据作为研究对象。将该模型的预测结果分别与粒子群优化算法和网格搜索算法优化的v-支持向量回归模型、遗传算法-支持向量回归模型进行比较。实验结果表明:遗传算法v-支持向量回归模型关于航迹经、纬度预测结果的均方误差、平均绝对百分比误差和平均绝对误差相比于其他模型最低,关于经度分别为4.29×10^(-7)(°)、4.50×10^(-4)和5.47×10^(-7)(°)2,关于纬度的分别为1.82×10^(-6)(°)、4.02×10^(-3)和1.08×10^(-3)(°)2。基于遗传算法-v支持向量回归模型的预测效果最好,预测误差波动最小。本文将遗传算法与v-支持向量回归相结合,为船舶轨迹预测模型的优化提供参考,也为海上智能交通提供思路。

黄浦江感潮河段潮位预报精度提升研究

黄浦江感潮河段水位波动以天文潮影响为主,同时也受上游径流、区间降雨和风暴潮等因素的影响,传统的调和分析方法难以考虑径流等非天文潮因素,在对黄浦江感潮河段进行潮位预报时总体精度偏低。本文在对传统调和分析方法预报误差频谱分析的基础上,提出了将传统调和分析模型和自回归模型相结合的预报方法,并将该方法应用于黄浦江感潮河段3个代表性站点(吴淞口、黄浦公园、米市渡)的潮位逐时预报中。结果表明:24 h潮位预报的均方根误差由原来的0.20 m左右降至0.10~0.14 m,预报精度显著提升。

基于GTWR的站域建成环境对城市轨道交通客流量的时空影响

轨道交通客流量影响因素是轨道交通方面研究的一个关注点,不同站点客流量的时空非平稳性被认为与站域建成环境有关。通过构建时空地理加权(geographically and temporally weighted regression,GTWR)模型,揭示了土地多样性、密度、站点属性3个方面因素在时间和空间维度上对天津市轨道交通客流量的影响。结果表明:相较于传统的地理加权(geographically weighted regression,GWR)模型和最小二乘法(ordinary least squares,OLS)模型,GTWR具有更好的拟合优度;公交站点密度对轨道交通客流产生促进作用,尤其在工作日的早晚高峰时段和中心城区位置;市中心的商业设施在工作日晚高峰吸引更多的地铁乘客,而在近郊区它们在早高峰吸引更多的地铁乘客;人口密度促进轨道交通的客流量;充足的停车场设施数量可以吸引更多的轨道交通乘客。

卷积神经网络的神华关键配件状态自动跟踪研究

为避免关键配件异常状态带来的神华铁路货车运行安全隐患,提出基于卷积神经网络的神华关键配件状态自动跟踪方法。神华关键配件状态图像作为卷积神经网络的输入数据,经过卷积和池化操作后获得神华关键配件状态检测结果,并将其作为图像第一帧的状态,然后利用核相关滤波训练获得的回归模型估计图像下一帧的状态,实现神华关键配件状态自动跟踪。实验结果表明:该方法能够获得较为完整、清晰的神华关键配件状态图像;不同神华关键配件状态检测的MCC值均在0.8以上,且能够在异常状态发生之前得到状态检测结果;各时刻的神华关键配件状态跟踪结果与实际结果完全相同。

基于支持向量回归的数控加工刀轨自动规划方法

传统的刀轨规划方法基于经验公式和手工计算,效率低下且难以保证加工精度,对此文章提出基于支持向量回归的数控加工刀轨自动规划方法。试验结果表明,基于支持向量回归的数控加工刀轨自动规划方法可提高效率与加工精度,具有较高的实用价值。

继电保护装置自动测试系统设计

分析了继电保护装置自动测试的现状和优势,针对继电保护装置研发过程中性能测试的要求和特点,介绍了一种通用的继电保护装置自动测试系统。该系统采用分布式体系结构和模块化设计思想,利用Python语言构造实时测试脚本,可实现保护装置的闭环自动测试和回归测试。通过改变配置,可以实现对常规站保护装置和数字化站保护装置的自动测试。对自动测试系统各模块的设计进行了详细说明,重点介绍了测试用例的实现方法。实际应用表明,该系统可提高研发过程中的测试质量及效率。

基于变化检测-CART决策树模式自动识别沙漠化信息

目前沙漠化遥感监测存在目视解译的局限性、数据源的约束性、遥感信息利用率低等问题。基于此,以民勤盆地为试验区,首先采用图像差值、最大值合成及二维最大类间方差等方法,检测1994年、2014年两期Landsat图像的变化像元,然后利用分类与回归树(CART)算法构建决策树,自动提取了2014年沙地信息,最后将变化检测结果与沙地信息进行空间叠置分析,并实现了沙漠化信息自动识别模式。研究表明,变化检测-CART决策树模式精度为89.43%~93.00%,在95%置信水平上其置信区间介于85.90%~98.00%,显然其精度具有较高可信度;该模式不仅能够充分利用丰富遥感信息而且可排除多余信息的干扰。可见,变化检测-CART决策树模式是识别沙漠化信息的有效方法之一,将对沙漠化防治工程具有重要应用价值。

遗传程序设计在统计建模中的应用

介绍一种利用遗传程序设计的方法来自动生成统计预测模型,并进行误差估计分析,改变过去只使用拟合曲线粗糙、预测结果不理想的几种传统固定统计预测模型的传统分析方法。通过对[1]和[2]的真实历史资料验证,结果表明,与传统的线性回归、指数回归、抛物线回归[3]三种方法对比,遗传程序设计建立的模型所预测的数据准确度明显要高。

基于Logistic回归建立霜自动判识模型

利用安徽省23个典型气象站2003-2017年观测数据,根据无霜日的气象要素阈值条件进行质量控制,在此基础上,构建各气象站基于Logistic回归的霜自动判识模型,并对模型的霜判识效果进行评估。结果表明:(1)通过气温、风速和降水量等气象要素阈值,能够有效判定出安徽各站当日无霜现象;(2)各气象站的霜判识模型均入选了温度、湿度和风速等相关要素作为判识因子,入选要素的时次多集中在4:00-8:00区间;(3)独立样本检验表明,基于Logistic回归的霜判识模型对安徽地区霜的平均判识准确率、命中率、漏判率、空判率和TS评分分别为89.0%、91.6%、8.4%、15.8%和78.2%,表明模型对安徽地区的霜具有较好的判识能力;(4)与Bayes判别模型对比发现,基于Logistic回归的霜判识模型在准确率、空判率和TS评分方面表现更优,可以使用Logistic回归模型实现霜的自动化判识。

基于回归算法的自动引导车跟踪控制

提出了一种跟踪控制效果好、适应性广的自动引导车跟踪方法。以自动引导车的动力学模型作为系统的被控对象 ,通过回归算法对系统以前的输入和输出跟踪误差信号进行运算来反复调整输入量 ,使得系统在经过一定次数的迭代以后 ,其实际输出趋于期望输出。仿真与试验结果均表明 ,装有此系统的自动引导车能够很好地跟踪给定的路径 ,验证了所提方法的正确性和有效性。

计算机联锁软件模拟自动测试系统的研究与实现

基于灰箱测试模型,实现非侵入式闭环联锁软件自动测试系统。系统测试分自动和手工两种方式。测试程序软件采用模块化设计。仿真程序界面采用虚拟面板技术,使操作直观、方便、简捷,增强了仿真能力。系统测试基于联锁特征数据,联锁表保证测试的连续进行。实际应用表明,本系统能够发现联锁软件故障,提高测试效率;有利于进行回归测试;可以降低人为的操作失误和对测试人员的技术要求,减少测试成本,满足联锁软件出厂测试和现场测试自动化的需要。

SAR图像目标方位角估计方法综述

目标方位角估计是进行SAR图像自动目标识别(ATR)的重要步骤。由于目标SAR图像对目标方位角的敏感性,在目标识别之前进行方位角估计能够提高系统的效率,增强系统的实时性。本文在分析SAR目标模型的基础上,对目标方位角估计算法进行了全面系统的研究,总结了目前存在的目标方位角估计方法的主要研究思路,即:提取目标主轴、提取目标主导边界、最小外接矩形获取,并介绍了有代表性的算法,指出了这些算法的优缺点,展望了今后的发展方向。