Multimodel Ensemble Forecast of Global Horizontal Irradiance at PV Power Stations Based on Dynamic Variable Weight

In the present study,multimodel ensemble forecast experiments of the global horizontal irradiance(GHI)were conducted using the dynamic variable weight technique.The study was based on the forecasts of four numerical models,namely,the China Meteorological Administration Wind Energy and Solar Energy Prediction System,the Mesoscale Weather Numerical Prediction System of China Meteorological Administration,the China Meteorological Administration Regional Mesoscale Numerical Prediction System-Guangdong,and the Weather Research and Forecasting Model-Solar,and observational data from four photovoltaic(PV)power stations in Yangjiang City,Guangdong Province.The results show that compared with those of the monthly optimal numerical model forecasts,the dynamic variable weight-based ensemble forecasts exhibited 0.97%-15.96%smaller values of the mean absolute error and 3.31%-18.40%lower values of the root mean square error(RMSE).However,the increase in the correlation coefficient was not obvious.Specifically,the multimodel ensemble mainly improved the performance of GHI forecasts below 700 W m^(-2),particularly below 400 W m^(-2),with RMSE reductions as high as 7.56%-28.28%.In contrast,the RMSE increased at GHI levels above 700 W m^(-2).As for the key period of PV power station output(02:00-07:00),the accuracy of GHI forecasts could be improved by the multimodel ensemble:the multimodel ensemble could effectively decrease the daily maximum absolute error(AE max)of GHI forecasts.Moreover,with increasing forecasting difficulty under cloudy conditions,the multimodel ensemble,which yields data closer to the actual observations,could simulate GHI fluctuations more accurately.

Bootstrapped Multi-Model Neural-Network Super-Ensembles for Wind Speed and Power Forecasting

The bootstrap resampling method is applied to an ensemble artificial neural network (ANN) approach (which combines machine learning with physical data obtained from a numerical weather prediction model) to provide a multi-ANN model super-ensemble for application to multi-step-ahead forecasting of wind speed and of the associated power generated from a wind turbine. A statistical combination of the individual forecasts from the various ANNs of the super-ensemble is used to construct the best deterministic forecast, as well as the prediction uncertainty interval associated with this forecast. The bootstrapped neural-network methodology is validated using measured wind speed and power data acquired from a wind turbine in an operational wind farm located in northern China.

基于生成对抗网络和对比学习的假新闻检测方法研究

社交媒体作为信息获取的主要途径,其假新闻问题日益严重。假新闻检测任务的重要挑战之一是确保模型能够及时响应新出现的事件,并在有限时间内完成检测任务,这要求模型具备高效的实时性和对新事件的快速适应能力,与此同时,多模态假新闻检测技术作为未来的重要发展方向也值得关注。针对上述挑战,提出了一种多模态假新闻检测模型ADSCL,利用卷积神经网络提取文本和图像的语义特征,并通过多层联合注意力机制进行融合。针对新事件的及时响应需求,引入生成对抗网络和对比学习,从大量数据中提取可转移特征,提高泛化能力。同时,通过对抗性训练增强模型鲁棒性。实验结果表明,ADSCL模型有效提升了假新闻检测能力,验证了多模态融合和对抗性方法在新闻检测任务上的优越性。

A Multimodel Ensemble-based Kalman Filter for the Retrieval of Soil Moisture Profiles

With the combination of three land surface models （LSMs） and the ensemble Kalman filter （EnKF）, a multimodel EnKF is proposed in which the multimodel background superensemble error covariance matrix is estimated by two different algorithms： the Simple Model Average （SMA） and the Weighted Average Method （WAM）. The two algorithms are tested and compared in terms of their abilities to retrieve the true soil moisture profile by respectively assimilating both synthetically-generated and actual near-surface soil moisture measurements. The results from the synthetic experiment show that the performances of the SMA and WAM algorithms were quite different. The SMA algorithm did not help to improve the estimates of soil moisture at the deep layers, although its performance was not the worst when compared with the results from the single-model EnKF. On the contrary, the results from the WAM algorithm were better than those from any single-model EnKF. The tested results from assimilating the field measurements show that the performance of the two multimodel EnKF algorithms was very stable compared with the single-model EnKF. Although comparisons could only be made at three shallow layers, on average, the performance of the WAM algorithm was still slightly better than that of the SMA algorithm. As a result, the WAM algorithm should be adopted to approximate the multimodel background superensemble error covariance and hence used to estimate soil moisture states at the relatively deep layers.

An Analysis of the Difference between the Multiple Linear Regression Approach and the Multimodel Ensemble Mean

An investigation of the difference in seasonal precipitation forecast skills between the multiple linear regression （MLR） ensemble and the simple multimodel ensemble mean （EM） was based on the forecast quality of individual models. The possible causes of difference in previous studies were analyzed. In order to make the simulation capability of studied regions relatively uniform, three regions with different temporal correlation coefficients were chosen for this study. Results show the causes resulting in the incapability of the MLR approach vary among different regions. In the Nifio3.4 region, strong co-linearity within individual models is generally the main reason. However, in the high latitude region, no significant co-linearity can be found in individual models, but the abilities of single models are so poor that it makes the MLR approach inappropriate for superensemble forecasts in this region. In addition, it is important to note that the use of various score measurements could result in some discrepancies when we compare the results derived from different multimodel ensemble approaches.

Multimodel Ensemble Forecasts for Precipitations in China in 1998

Different multimodel ensemble methods are used to forecast precipitations in China, 1998, and their forecast skills are compared with those of individual models. Datasets were obtained from monthly simulations of eight models during the period of January 1979 to December 1998 from the “Climate of the 20th Century Experiment” （20C3M） for the Fourth IPCC Assessment Report. Climate Research Unit （CRU） data were chosen for the observation analysis field. Root mean square （RMS） error and correlation coeffi-cients （R） are used to measure the forecast skills. In addition, superensemble forecasts based on different input data and weights are analyzed. Results show that for original data, superensemble forecasting based on multiple linear regression （MLR） performs best. However, for bias-corrected data, the superensemble based on singular value decomposition （SVD） produces a lower RMS error and a higher R than in the MLR superensemble. It is an interesting result that the SVD superensemble based on bias-corrected data performs better than the MLR superensemble, but that the SVD superensemble based on original data is inferior to the corresponding MLR superensemble. In addition, weights calculated by different data formats are shown to affect the forecast skills of the superensembles. In comparison with the MLR superensemble, a slightly significant effect is present in the SVD superensemble. However, both the SVD and MLR superensembles based on different weight formats outperform the ensemble mean of bias-corrected data.

MULTIMODEL CONSENSUS FORECASTING OF LOW TEMPERATURE AND ICY WEATHER OVER CENTRAL AND SOUTHERN CHINA IN EARLY 2008

Based on the daily mean temperature and 24-h accumulated total precipitation over central and southern China, the features and the possible causes of the extreme weather events with low temperature and icing conditions,which occurred in the southern part of China during early 2008, are investigated in this study. In addition, multimodel consensus forecasting experiments are conducted by using the ensemble forecasts of ECMWF, JMA, NCEP and CMA taken from the TIGGE archives. Results show that more than a third of the stations in the southern part of China were covered by the extremely abundant precipitation with a 50-a return period, and extremely low temperature with a 50-a return period occurred in the Guizhou and western Hunan province as well. For the 24- to 216-h surface temperature forecasts, the bias-removed multimodel ensemble mean with running training period(R-BREM) has the highest forecast skill of all individual models and multimodel consensus techniques. Taking the RMSEs of the ECMWF 96-h forecasts as the criterion, the forecast time of the surface temperature may be prolonged to 192 h over the southeastern coast of China by using the R-BREM technique. For the sprinkle forecasts over central and southern China, the R-BREM technique has the best performance in terms of threat scores(TS) for the 24- to 192-h forecasts except for the 72-h forecasts among all individual models and multimodel consensus techniques. For the moderate rain, the forecast skill of the R-BREM technique is superior to those of individual models and multimodel ensemble mean.

Improving Seasonal Climate Forecasts over Various Regions of Africa Using the Multimodel Superensemble Approach

Improvements that can be attained in seasonal climate predictions in various parts of Africa using the multimodel supersensemble scheme are presented in this study. The synthetic superensemble (SSE) used follows the approach originally developed at Florida State University (FSU). The technique takes more advantage of the skill in the climate forecast data sets from atmosphere-ocean general circulation models running at many centres worldwide including the WMO global producing centers (GPCs). The module used in this work drew data sets from the Four versions of FSU coupled model system, seven models from the DEMETER project which is the forerun to the current European Ensembles Forecast System, the NCAR Model, and the Predictive Ocean Atmosphere Model for Australia (POAMA), all making a set of 13 individual models. An archive consisting of monthly simulations of precipitation was available over all the 5 regions of Africa, namely Eastern, Central, Northern, Southern, and Western Africa. The results showed that the SSE forecast for precipitation carries a higher skill compared to each of the member models and the ensemble mean. Relative to the ensemble mean (EM), the SSE provides an improvement of 18% in simulation of season cycle of precipitation climatology. In Eastern Africa, during December-February season, a north-south gradient of precipitation prevails between Tropical East Africa and the sector of the region towards Southern Africa. This regional scale climate pattern is a direct influence of the Intertropical Convergence Zone (ITZC) across the African continent during this time of the year. The SSE emerges with superior skill scores such as lowest root mean square error above the EM and the member models, for example in the prediction of spatial location and precipitation magnitudes that characterize the see-saw precipitation pattern in Eastern Africa. In all parts of Africa, and especially Eastern Africa where seasonal precipitation variability is a frequent cause huge human suffering due to droughts and famine, the multimodel superensemble and its subsequent improvements will always provide a forecast that outweighs the best Atmosphere-Ocean Climate Model. This approach and results herein imply that climate services centres worldwide and Africa in particular can make more objective use of model forecast data sets provided by global producing centres (GPCs) for consensus climate outlooks.

Segmentation and texture analysis with multimodel inference for the automatic detection of exudates in early diabetic retinopathy

Diabetic retinopathy (DR) is an eye disease caused by the increase of insulin in blood and may cause blindness if not treated at an early stage. Exudates are the primary sign of DR. Currently there is no fully automated method to detect exudates in the literature and it would be useful in large scale screening if fully automatic method is available. In this paper we developed a novel method to detect exudates that based on interactions between texture analysis and segmentation with mathematical morphological technique by using multimodel inference. The texture analysis involves three components: they are statistical texture analysis, high order spectra analysis, and fractal analysis. The performance of the proposed method is assessed by the sensitivity, specificity and accuracy using the public data DIARETDB1. Our results show that the sensitivity, specificity and accuracy are 95.7%, 97.6% and 98.7% (SE = 0.01), respectively. It is shown that the proposed method can be run automatically and also improve the accuracy of exudates detection significantly over most of the previous methods.

Multimodel Approach for Intelligent Control and Applications

The use of the multimodel approach in the modelling, analysis and control of non-linear complex and/or ill-defined systems was advocated by many researchers. This approach supposes the definition of a set of local models valid in a given region or domain. Different strategies exist in the literature and are generally based on a partitioning of the non-linear system’s full range of operation into multiple smaller operating regimes each of which is associated with a locally valid model or controller. However, most of these strategies, which suppose the determination of these local models as well as their validity domain, remain arbitrary and are generally fixed thanks to a certain a priori knowledge of the system whatever its order. Recently, we have proposed a new approach to derive a multimodel basis which allows us to limit the number of models in the basis to almost four models. Meanwhile, the transition problem between the different models, which may use either a simple commutation or a fusion technique, remains still arise. In this plenary talk, a fuzzy fusion technique is presented and has the following main advantages: （1） use of a fuzzy partitioning in order to determine the validity of each model which enhances the robustness of the solution; 2 introduction, besides the four extreme models, of another model, called average model, determined as an average of the boundary models.

华东地区地面和高空风场的多模式集成精细化预报研究

基于欧洲中期天气预报中心(European Centre for Medium-Range Weather Forecasts,ECMWF)、中国国家气象中心业务运行的中尺度数值预报系统(Global/Regional Assimilation and Prediction Enhanced System Meso,GRAPES-Meso)、美国国家环境预报中心(National Centers for Environmental Prediction,NCEP)的全球预报系统(Global Forecast System,GFS)、GRAPES全球预报系统(GRAPES-GFS)4个模式风场预报资料,利用双线性、反距离加权、三次样条、克里格等插值方法对华东及周边地区(110°~130°E,20°~40°N)2020年1—4月逐日地面和高空风0~72 h集合预报资料进行降尺度处理,得到满足机场及终端区气象保障的精细化风场预报。此外,还对精细化风场预报做多模式集成。结果表明,对于风场的精细化格点预报,反距离加权插值方法误差最小,为最优水平插值方法。基于扩展复卡尔曼滤波的多模式集成(Augmented Complex Extended Kalman Filter,ACEKF)可进一步减小风场预报的误差。对华东地区上海、青岛和厦门3个机场地面和高空风的多模式集成风场精细化预报的分析表明,ACEKF多模式集成预报不但均方根误差较BREM、ECMWF和GRAPES-GFS的预报误差小,且随高度变化也不如单模式预报的大,其预报性能更为稳定。

An Intelligent Adaptive Fuzzy PID Controller Based on Multimodel Control Approach

A novel intelligent adaptive fuzzy PHD controller based on multimodel control approach is presented in this paper.It can improve the system performance of the dynamic time- varying system at various operating conditions.The fuzzy PHD controller is implemented by combining a fuzzy PI with a fuzzy PD controller in a parallel structure. The parameters of the fuzzy PHD controller are linked, via analytical derivation, to the gains of the linear PID controller. The sum of error square is used as performance criterion to locate the model that best reresents the process among the multiple models, The desired control output to drive the process along the desired path is generated only by modifying the output scale factots GU_I and GU_D of the fuzzy PID controller, Among the prescribed models, the control signal of the nearestmmodel to the system is applied. The system can be driven to its original trajectory because of the robustness of the fuzzy PID controller, Computer simulation results show that the

Multimodal feature fusion based on object relation for video captioning

Video captioning aims at automatically generating a natural language caption to describe the content of a video.However,most of the existing methods in the video captioning task ignore the relationship between objects in the video and the correlation between multimodal features,and they also ignore the effect of caption length on the task.This study proposes a novel video captioning framework(ORMF)based on the object relation graph and multimodal feature fusion.ORMF uses the similarity and Spatio-temporal relationship of objects in video to construct object relation features graph and introduce graph convolution network(GCN)to encode the object relation.At the same time,ORMF also constructs a multimodal features fusion network to learn the relationship between different modal features.The multimodal feature fusion network is used to fuse the features of different modals.Furthermore,the proposed model calculates the length loss of the caption,making the caption get richer information.The experimental results on two public datasets(Microsoft video captioning corpus[MSVD]and Microsoft research-video to text[MSR-VTT])demonstrate the effectiveness of our method.

Dynamical Predictability of Leading Interannual Variability Modes of the Asian-Australian Monsoon in Climate Models

The dynamical prediction of the Asian-Australian monsoon(AAM)has been an important and long-standing issue in climate science.In this study,the predictability of the first two leading modes of the AAM is studied using retrospective prediction datasets from the seasonal forecasting models in four operational centers worldwide.Results show that the model predictability of the leading AAM modes is sensitive to how they are defined in different seasonal sequences,especially for the second mode.The first AAM mode,from various seasonal sequences,coincides with the El Niño phase transition in the eastern-central Pacific.The second mode,initialized from boreal summer and autumn,leads El Niño by about one year but can exist during the decay phase of El Niño when initialized from boreal winter and spring.Our findings hint that ENSO,as an early signal,is conducive to better performance of model predictions in capturing the spatiotemporal variations of the leading AAM modes.Still,the persistence barrier of ENSO in spring leads to poor forecasting skills of spatial features.The multimodel ensemble(MME)mean shows some advantage in capturing the spatiotemporal variations of the AAM modes but does not provide a significant improvement in predicting its temporal features compared to the best individual models in predicting its temporal features.The BCC_CSM1.1M shows promising skill in predicting the two AAM indices associated with two leading AAM modes.The predictability demonstrated in this study is potentially useful for AAM prediction in operational and climate services.

帕瑞昔布钠加入多模式镇痛对胸腔镜辅助开胸术后吗啡用量的影响

目的观察帕瑞昔布钠对胸腔镜辅助开胸术镇痛效果和术后吗啡静脉自控镇痛(PCA)用量的影响。方法选择胸腔镜辅助开胸手术患者100例,随机分为5组(P1,P2,P3,P4,P5组),每组20例,行双盲对照观察。5组病人均在切皮前给予帕瑞昔布钠40 mg静脉注射,关胸前胸腔内肋间神经阻滞以切口为中心上下三个肋间注射利罗合剂4～5 ml,术后5组均给予吗啡静脉PCA泵。P1组为对照组(生理盐水200 ml);P2组(吗啡5 mg+生理盐水=200 ml);P3组(吗啡10 mg+生理盐水=200 ml);P4组(吗啡15 mg+生理盐水=200 ml);P5组(吗啡20 mg+生理盐水=200 ml)。术后每24 h静脉注射帕瑞昔布钠40 mg。观察:静息、咳嗽视觉模拟评分(VAS);患者呼吸功能评价(RR、SpO2、EtCO2、VT)。记录PCA泵开启后1、2、4、8、12、24、36、48 h时点上述各项指标,在各时段内记录PCA泵的实际按压次数(D1)及有效按压次数(D2)。结果 5组患者在术后24 h内均无呼吸抑制,在术后0～6 h静息VAS评分无显著性差异;P1组的8～24 h静息、咳嗽VAS评分明显高于P2、P3、P4、P5组(P<0.05),P2、P3、P4、P5组间比较无显著性差异(P>0.05);36～48 h各组静息、咳嗽VAS评分无显著差异。P1、P2组的D1/D2在4～24 h与其他各组比较有显著性差异(P<0.05),P3～P5组之间无显著性差异。结论帕瑞昔布钠加入胸腔镜辅助开胸术后多模式镇痛,可以减少吗啡用量,镇痛效果良好,对开胸手术病人呼吸影响小,有利于病人排痰。

北半球中纬度地区地面气温的超级集合预报

基于TIGGE资料中的ECMWF、JMA、NCEP和UKMO四个中心2007年6月1日-8月31日北半球中纬度地区地面气温24～168h集合预报资料，分别利用固定训练期超级集合（SUP，Superensemble）和滑动训练期超级集合（R—SUP，Running Training Period Superensemble）对2007年8月8—31日预报期24d进行超级集合预报试验。采用均方根误差对预报结果进行检验评估，比较了两种超级集合方法与最好的单个中心模式预报、多模式集合平均的预报效果。结果表明，SUP预报有效降低了预报误差，24～144h的预报效果优于多模式集合平均（EMN，Ensemble Mean）和最好的单个中心预报，168h的预报效果略差于EMN。R－SUP预报进一步改善了预报效果。对于24～168h的预报，R－SUP预报效果都要优于EMN。尤其对于168h的预报，R－SUP改进了预报效果，优于EMN。

隔离小生境遗传算法研究

小生境 (niche)技术的引入 ,提高了遗传算法处理多峰函数 (m ultimodel function)优化问题的能力 .本文提出了基于隔离 (Isolation)机制的小生境技术 .隔离小生境技术具有生物学基础 ,不仅能够有效地保证群体中解的多样性 ,而且具有很强的引导进化能力 .计算机模拟旅行商推销问题 (TSP)的结果表明 。

基于TIGGE资料的西太平洋热带气旋多模式集成预报方法比较

基于TIGGE资料中CMA、ECMWF、JMA和NCEP四中心2010、2011和2012年3年的资料,采用集合平均(EMN)、加权集合平均(WEMN)、消除偏差集合平均(BREM)和加权消除偏差集合平均(即超级集合,SUP)四种方法,对西太平洋地区热带气旋路径与中心气压进行时效为24、48、72、96和120 h的多模式集成预报,评估了不同单中心预报结果,并分析了不同多模式集成预报方法的特点,对比了不同多模式集成方法的效果。结果表明,对于热带气旋路径和中心气压预报,各中心预报技巧不同,其中3年的CMA预报效果均较差,2010、2011年的ECMWF预报和2012年的NCEP预报效果最优;总体上几种多模式集成方法在120 h预报时效内均优于单模式预报,其中EMN作为一种最简单的集成预报方法,具有一定的局限性,而WEMN由于给不同单模式预报赋予了权重,因此相对于EMN能够得到更好的多模式集成预报结果;BREM方案由于消除了模式预报中的系统性偏差,因此集成预报效果也优于EMN;由于去除了模式预报偏差,同时给不同模式赋予了权重,SUP方案得到的集成预报效果最优。

降水集合预报集成方法研究

基于TIGGE(the THORPEX Interactive Grand Global Ensemble)资料,对中国气象局(CMA)、欧洲中期天气预报中心(ECMWF)、美国国家环境预报中心(NCEP)和日本气象厅(JMA)的集合数值预报结果进行降水集成。采用算术平均法、TS评分集成法和BS评分集成法在我国东南地区进行降水集成,对比分析结果表明:基于TS评分的多模式降水集成无论在分区降水评分中,还是在东南地区的台风型降水和非台风型降水实例中,都有效地改进了大雨以上的降水预报效果;基于BS评分的集成方法和算数平均集成法预报效果次之。东南地区5个子区域的降水集成试验结果表明:各子区域基于TS评分集成后降水的平均绝对误差普遍小于基于BS评分后的降水平均绝对误差。广东东南和浙江北部区域基于TS集成后的降水TS评分值最优,浙闽沿海和广东西北部区域基于TS集成后的降水TS评分次之,处于中上水平。基于算术平均集成和BS集成的降水的TS评分值只有在广东东南区域表现出较好的效果。

基于TIGGE资料的地面气温和降水的多模式集成预报

利用TIGGE资料集下中国气象局(CMA)、欧洲中期天气预报中心(ECMWF)、日本气象厅(JMA)、美国国家环境预报中心(NCEP)和英国气象局(UKMO)5个中心集合预报结果,对多模式集成预报方法进行讨论。结果表明,多模式集成方法的预报效果优于单个中心的预报,但对于不同预报要素多模式集成方法的适用性存在差异。滑动训练期超级集合(R-SUP)对北半球地面气温的改进效果最优,但此方法对降水场的改进效果并不理想。在北半球中低纬24h累积降水的回报试验中,消除偏差(BREM)的结果优于单个中心的预报,且此方法预报结果稳定。进一步利用滑动训练期消除偏差(R-BREM)集合平均对2008年1月中国南方极端雨雪冰冻过程进行多模式集成预报试验,结果表明,在固定误差范围内,R-BREM将中国南方大部分地区的地面气温预报时效由最优数值预报中心的96h延长至192h,且除个别时效外,小雨、中雨的TS评分得到明显提高。