Weather Prediction With Multiclass Support Vector Machines in the Fault Detection of Photovoltaic System

Since the efficiency of photovoltaic(PV) power is closely related to the weather,many PV enterprises install weather instruments to monitor the working state of the PV power system.With the development of the soft measurement technology,the instrumental method seems obsolete and involves high cost.This paper proposes a novel method for predicting the types of weather based on the PV power data and partial meteorological data.By this method,the weather types are deduced by data analysis,instead of weather instrument A better fault detection is obtained by using the support vector machines(SVM) and comparing the predicted and the actual weather.The model of the weather prediction is established by a direct SVM for training multiclass predictors.Although SVM is suitable for classification,the classified results depend on the type of the kernel,the parameters of the kernel,and the soft margin coefficient,which are difficult to choose.In this paper,these parameters are optimized by particle swarm optimization(PSO) algorithm in anticipation of good prediction results can be achieved.Prediction results show that this method is feasible and effective.

A Novel Uniplanar Multi-Electrode Capacitive Sensor for In-Situ Weathering Damage Detection of Nonmetallic Materials

A uniplanar capacitive sensor with 5-electrodes on one plane substrate and a large reflector electrode,was designed to get the corresponding capacitance information for weathering damage detection of non-metallic materials exposed to a service environment.A 2-D finite-element method was employed to simulate the electric potential distribution and capacitance measurements for the sensor.2 marble slabs,one was healthy and the other was notched,were experimentally detected.Both the simulation and the preliminary experimental results show that the measured capacitances decrease after weathering damage occurs in nonmetallic material.The reflector can enlarge the sensitive depth.The weathering assessment of nonmetallic materials can be done by processing the measured capacitances.The proposed approach can effectively detect the weathering damage of nonmetallic material and can be practically used for in-situ weathering damage evaluation.

Deep convolutional neural network for meteorology target detection in airborne weather radar images

Considering the problem that the scattering echo images of airborne Doppler weather radar are often reduced by ground clutters,the accuracy and confidence of meteorology target detection are reduced.In this paper,a deep convolutional neural network(DCNN)is proposed for meteorology target detection and ground clutter suppression with a large collection of airborne weather radar images as network input.For each weather radar image,the corresponding digital elevation model(DEM)image is extracted on basis of the radar antenna scan-ning parameters and plane position,and is further fed to the net-work as a supplement for ground clutter suppression.The fea-tures of actual meteorology targets are learned in each bottle-neck module of the proposed network and convolved into deeper iterations in the forward propagation process.Then the network parameters are updated by the back propagation itera-tion of the training error.Experimental results on the real mea-sured images show that our proposed DCNN outperforms the counterparts in terms of six evaluation factors.Meanwhile,the network outputs are in good agreement with the expected mete-orology detection results(labels).It is demonstrated that the pro-posed network would have a promising meteorology observa-tion application with minimal effort on network variables or parameter changes.

Study on Query System of the Historical High-altitude Ground Weather Chart

We introduced query system of the historical high-altitude surface weather chart. Historical high-altitude surface data were converted to Grads data format. Grads as generation tool of the image, ASP was used to compile WEB page. By B/S mode, only user submitted necessary conditions for the image to server by client browser, historical high-altitude surface weather chart at corresponding time and height could be ob- tained. Without any procedures and related data, only needed client browser, user could use image conveniently to a large extent.

Assimilation of FY-3D MWTS-Ⅱ Radiance with 3D Precipitation Detection and the Impacts on Typhoon Forecasts

Precipitation detection is an essential step in radiance assimilation because the uncertainties in precipitation would affect the radiative transfer calculation and observation errors.The traditional precipitation detection method for microwave only detects clouds and precipitation horizontally,without considering the three-dimensional distribution of clouds.Extending precipitation detection from 2D to 3D is expected to bring more useful information to the data assimilation without using the all-sky approach.In this study,the 3D precipitation detection method is adopted to assimilate Microwave Temperature Sounder-2(MWTS-Ⅱ)onboard the Fengyun-3D,which can dynamically detect the channels above precipitating clouds by considering the near-real-time cloud parameters.Cycling data assimilation and forecasting experiments for Typhoons Lekima(2019)and Mitag(2019)are carried out.Compared with the control experiment,the quantity of assimilated data with the 3D precipitation detection increases by approximately 23%.The quality of the additional MWTS-Ⅱradiance data is close to the clear-sky data.The case studies show that the average root-mean-square errors(RMSE)of prognostic variables are reduced by 1.7%in the upper troposphere,leading to an average reduction of4.53%in typhoon track forecasts.The detailed diagnoses of Typhoon Lekima(2019)further show that the additional MWTS-Ⅱradiances brought by the 3D precipitation detection facilitate portraying a more reasonable circulation situation,thus providing more precise structures.This paper preliminarily proves that 3D precipitation detection has potential added value for increasing satellite data utilization and improving typhoon forecasts.

A Precipitation Detection Method for MWTS-Ⅱ Radiance Assimilation in Typhoon Simulation

FY-3C Microwave Temperature SounderⅡ(MWTS-Ⅱ)lacks observations at 23.8 GHz,31 GHz and 89 GHz,making it difficult to remove the data contaminated by precipitation in assimilation.In this paper,a fast forward operator based on the Community Radiative Transfer Model(CRTM)was used to analyze the relationship between the observation minus background simulation(O-B)and the cloud fractions in different MWTS-Ⅱchannels.In addition,based on the community Gridpoint Statistical Interpolation(GSI)system,the radiation brightness temperature of the MWTS-Ⅱwas assimilated in the regional Numerical Weather Prediction(NWP)model.In the process of assimilation,Visible and Infrared Radiometer(VIRR)cloud detection products were matched to MWTS-Ⅱpixels for precipitation detection.For typhoon No.18 in 2014,impact tests of MWTS-Ⅱdata assimilation was carried out.The results show that,though the bias observation minus analysis(O-A)of assimilated data can be reduced by quality control only with|O-B|<3 K;however,the O-A becomes much smaller while the precipitation detection is performed with Fvirr<0.9(VIRR cloud fraction threshold of 0.9).Besides,the change of the environmental field around the typhoon is more conducive to make the simulated track closer to the observation.The 72-hour typhoon track simulation error also shows that,after the precipitation detection,the error of simulated typhoon track is significantly reduced,which reflects the validity of a precipitation detection method based on a double criterion of|O-B|<3 K and Fvirr<0.9.

Detecting Climate Change in Using Extreme Data from Two Surface Weather Stations: Case Study Valle of Comitan and La Esperanza, Chiapas, Mexico

The study area is located between the cities of Comitan (16°10'43"N and 92°04'20''W) a city with 150,000 inhabitants and La Esperanza (16°9'15''N and 91°52'5''W) a town with 3000 inhabitants. Both weather stations are 30 km from each other in the Chiapas State, México. 54 years of daily records of the series of maximum (tmax) and minimum temperatures (tmin) of the weather station 07205 Comitan that is on top of a house and 30 years of daily records of the weather station 07374 La Esperanza were analyzed. The objective is to analyze the evidence of climate change in the Comitan valley. 2.07% and 19.04% of missing data were filled, respectively, with the WS method. In order to verify homogeneity three methods were used: Standard Normal Homogeneity Test (SNHT), the Von Neumann method and the Buishand method. The heterogeneous series were homogenized using climatol. The trends of tmax and tmin for both weather stations were analyzed by simple linear regression, Sperman’s rho and Mann-Kendall tests. The Mann-Kendal test method confirmed the warming trend at the Comitan station for both variables with ZMK statistic values equal to 1.57 (statistically not significant) and 4.64 (statistically significant). However, for the Esperanza station, it determined a cooling trend for tmin and a slight non-significant warming for tmax with a ZMK statistic of -2.27 (statistically significant) and 1.16 (statistically not significant), for a significance level α = 0.05.

Analysis of Detected Avalanches Using Meteorological Data of Nearby Monitoring Stations in Ischgl, Austria

A set of detected avalanches from January to April 2012 on a hillside southeast of lschgl, Austria is given. The avalanches are off-the-cut or caused by blast. The meteorological data of two monitoring stations nearby the hillside are taken for analysing the weather situation. The meteorological parameters air temperature, wind intensity and wind speed, relative humidity, precipitation and snow depth are investigated for similarities short before and during an avalanche. The avalanches are grouped into three categories and meteorological characteristics are found for each category. Thereby the avalanche hazard for the observed hillside is better assessed and an infrastructure safety by avalanche control due to concerted avalanche blasts is more effective. The result of the analysis shows three kinds of hazard weather conditions, which increase the avalanche hazard： warm air temperatures cause a settlement of the snow pack, but in the beginning of the process a weakening in the snow pack happens. Rapidly decreasing of the air temperature cause cracks in the snow pack and the combination of fresh snow and strong wind speed leads to accumulation of snow on sheltered slopes.

月球空间天气探测与研究进展

稳定的天体地质构造、与地球合适的距离、无稠密大气和全球磁场,使得月球成为日地月空间天气监测和研究的天然优良实验室。回顾了国内外月球空间天气探测和研究进展,梳理了月球空间天气关键科学问题及预报应用问题,提出了建立月球空间天气监测站的构想,明确了监测站具体科学目标,并介绍了监测站三大系统构成及相互关系,包括监测系统、科学研究系统以及建模与预报系统。建立月球空间天气监测站对研究太阳爆发活动、日地月耦合关系以及月球局地环境的变化具有重要的科学意义,也有利于促进月球空间天气建模和预报技术的发展,提升未来月球科研探索任务空间环境保障能力。

中国地面自动气象站小时降水质量控制方法

高时空分辨率自动气象站降水观测作为重要数据来源,已被广泛应用于强对流天气监测、模式评估、预报复盘等研究工作。仪器故障、特殊天气条件下观测设备的局限性等因素是自动气象站降水数据不确定性的主要来源,这些问题在无人值守气象站尤为突出。该研究基于2021—2023年中国自动气象站实时观测降水量数据、高时空分辨率雷达数据和高灵敏性降水类天气现象数据,发展适应于中国自动气象站小时降水数据的多源数据协同质量控制方法(multi-source data collaborative quality control,MDC)。通过综合定量指标与典型个例分析,对MDC的应用效果进行全面评估。结果显示:MDC判识正确率为99.92%,错误数据命中率较现行业务提升39.3%。基于多源降水观测数据时空一致性,MDC显著提升了晴空降水、融雪性降水和虚假零值降水等异常数据的甄别能力,有效弥补了传统方法的不足。

基于置信规则库气象站故障检测专家系统

针对气象站故障检测的需求,研制了基于置信规则库的气象站故障检测专家系统。专家系统利用前期建立起来的气象站的气象数据库,根据当前实时气象数据,通过改进的置信规则库模型结合专家经验知识自动完成数据分析和处理、是否有故障判断并对故障程度进行等级划分。工作人员在进行现场故障检测之前,只需通过气象站传感器故障专家系统对异常传感器进行辅助诊断,确定故障等级后依据故障等级携带相应检修设备或需更换的零部件进入现场实施检查和维修。实验测试表明系统对各个设备的总体平均检测准确率从69.8%提高到85.7%,极大提升了检测的效率和准确性。

云冈石窟风化检测研究进展

云冈石窟,作为我国规模最大的古代石窟群,是文化遗产的重要组成部分。由于年代久远,许多石质文物都遭受着严重的风化病害侵蚀,对其风化程度进行检测已经成为石质文物保护中的一项紧迫任务。随着检测技术的不断发展,石质文物风化的检测方法也越来越多,并且每种检测方法都有其侧重的检测目的和优势。本研究综述近几十年内曾使用过的重点检测方法,并分析每种检测方法的原理、优势和局限性,为云冈石窟风化检测和保护提供参考。

垂直大气湍流影响下的天气雷达回波信号优化检测方法

垂直大气湍流导致回波信号在大气中扩散,使回波信号质量较差,极易被噪声所淹没,因此,提出垂直大气湍流影响下的天气雷达回波信号优化检测方法。通过分析垂直大气湍流对天气雷达回波信号传输特性的影响,匹配滤波优化算法,改善垂直大气湍流影响导致质量下降的回波信号,利用模糊逻辑方法提取回波特征参数并构建对应隶属函数,设定阈值,获取天气雷达回波信号检测结果。实验结果表明:该方法可有效检测不同仰角的天气雷达回波信号,回波信号波形均匀且波动幅度小,检测效果较好。

低空复杂风场全天候雷达精细探测技术

低空复杂风场通常指距离地面高度600 m以下,局地风速、风向快速变化的风场.其全天候大范围精细探测是雷达科学、气象学等领域的长期难题.低空复杂风场的全天候雷达精细探测涉及传感器设计、三维风场反演和全天候融合等关键技术.本文从激光雷达探测、微波雷达探测、全天候融合及特征危害提取等方面对风场探测的技术内涵及研究现状进行了梳理,并分析了发展趋势,为低空复杂风场全天候精细探测技术发展与应用提供参考.

基于改进YOLOv8s的恶劣天气车辆行人检测方法

针对恶劣天气条件下摄像头捕获图像时存在图像模糊以及光照分布不均等问题,导致了场景对比度的下降,进而增加了在图像中区分检测目标与背景的难度。为了提高在恶劣天气环境下车辆和行人的检测能力,本文提出了一种改进的YOLOv8s算法。首先,本文在YOLOv8s算法的基础上,利用可扩张残差结构对主干网络中的C2F模块进行了优化,增强了模型对环境变化的适应能力。同时,在主干网络的SPPF模块前置层引入了高效多尺度注意力机制,该机制能够更有效地捕获图像中丰富多变的多尺度特征。其次,针对YOLOv8s算法的检测头部进行了重新设计,在保持模型准确性的前提下,降低了模型的复杂度。最后,引入Wise-IoU改进YOLOv8s算法的回归损失函数,提高了算法的收敛速度和检测精度。实验结果表明,改进的YOLOv8s算法在恶劣天气条件下对车辆和行人检测的平均精度均值达到91.41%,相比原始算法提升了2.56%,同时模型参数量减少了8%,计算量降低了4.9 GFLOPs。相比于其他主流的目标检测算法,改进后的YOLOv8s算法在保证了实时性能的同时,满足了恶劣天气条件下的车辆和行人检测需求。

结合循环提取器与自蒸馏的目标检测方法

在深度学习时代,目标检测方法不断发展,且在良好的视觉环境中已经具有较高的水平。但是,在特殊天气下,常规的目标检测方法的检测性能大幅下降,甚至失效,而特殊天气环境的行车安全一直是社会广泛关注的问题。为了解决上述问题,该文主要设计了一个目标检测器的模型,即引入循环解纠缠、自蒸馏方法的改进YOLO模型。在循环解纠缠模块,从输入图像中循环提取域不变特征,通过循环操作,可以在不依赖域相关注释的情况下,提高图像域特征和域不变特征的解缠能力;在自蒸馏模块,以提取到的域不变特征为教师对象,进一步提高泛化能力。并且该检测器在只有一个源域进行训练的情况下,面对许多未曾训练过的目标域上仍然表现良好,提高了检测器在未知域的鲁棒性。实验验证了模型在各种天气下城市场景目标检测的效果,实验数据表明,该方法优于基线约8百分点,相比基线方法获得了性能提升。

基于3种方法的福建天气雷达探测环境评估与分析

【目的】我国新一代天气雷达组网探测仍然存在较大盲区,特别是对低层大气的覆盖情况缺少评估。该文旨在研究福建省雷达对低层大气的探测情况,以辅助补盲雷达站址选择。【方法】利用SRTM高分辨率地形数据,采用3种方法,对福建省8部SA天气雷达0.5~3.4°仰角的探测环境,从单站范围、雷达探测重叠度、组网覆盖范围等方面进行分析,其中方法1、方法2、方法3分别基于射线海拔高度不同、射线与地形高度差不同、馈源以上高度不同,对一定高度区间的雷达覆盖情况进行评估。【结果】从业务角度出发评估低层大气探测环境,方法1和方法2的评估效果更优,以方法2为例,分析在盲区新建雷达的补盲效果,结果表明盲区主要位于闽中、闽西南和闽北地区。【结论】补盲雷达若位于靠近福建省界的闽西南区域,能最大程度新增0~2 km高度雷达波束覆盖面积,经计算得出新增面积可达15 000 km2以上。

基于机器视觉的雨雾天驾驶辅助系统设计

随着私家车数量日益增多,雨雾天交通安全问题成为了亟待解决的难题;在嵌入式硬件资源有限的情况下,为驾驶用户设计了基于机器视觉的雨雾天语音辅助驾驶系统;系统结合了湿度传感器、轻量化去雾神经网络AOD-NET和目标检测模型YOLOv5n;在目标检测模型YOLOv5n上,利用K-means++算法重新设计锚框,选取较优的骨干网络并利用模型剪枝进一步压缩模型大小;实验结果表明,改进的模型在Jetsonnano上的FPS达到了17.78,最终mAP在人工加雾、分辨率变化的TT100K (Tsinghua-Tencent 100K)数据集到达了65.8%,满足了正常天气与雨雾天气下的驾驶辅助实际应用。

基于LSTM预测模型的应用性能异常检测

目前高性能计算系统规模和复杂性不断增加,应用软件作业性能异常的原因变得更加复杂多样,传统的针对基于监控数据进行人工分析的方法存在效率低下和过分依赖分析人员经验的问题。提出一种基于长短期记忆网络(LSTM)的性能异常检测方法。以天气预报模式WRF为研究对象,首先从历史作业数据中学习出正常性能数据的变化情况,然后通过引入boxplot方法对LSTM模型预测值与实际观测值之间的残差进行统计分析,并将大于下四分位的数据判定为异常,从而实现应用软件作业性能异常的检测。实验结果表明,上述方法不仅可以较好地检测出性能的异常,而且能适用于多种不同类型的数据集。

硬质探空落球数传系统天线设计方法

针对硬质探空落球高空惯性运动工况及落球内部有限安装空间下的无线传输需求,给出了适用于临近空间气象探测落球数传系统的天线设计方法。该方法采用中心对称曲折线振子天线技术方案,解决了在狭小空间内实现全向天线覆盖的难题。实验结果表明,该方法有效解决了落球在空中下落过程中由于姿态变化引起的信号起伏大从而影响无线信号可靠传输的问题。