Abstract:This paper studies the auxiliary model based stochastic gradient(AM-SG) algorithm,the auxiliary model based multi-innovation stochastic gradient(AM-MISG) algorithm,the auxiliary model based recursive least squares(AM-RLS) algorithm and the auxiliary model based multi-innovation least squares algorithm for output-error systems and output-error autoregressive moving average(OEARMA) systems(namely,Box-Jenkins systems),the modified AM-SG algorithm,the forgetting factor AM-SG algorithm,the interval-varying AM-SG algorithm,the interval-varying AM-MISG algorithm and the interval-varying AM-RLS algorithm for output-error systems,and presents the filtering based auxiliary model generalized stochastic gradient algorithm and the filtering based multi-innovation generalized least squares algorithm for output-error autoregressive systems(namely,OEAR systems).

Abstract:The key aim of this serial is to study various stochastic models in finance with emphasis on the Monte Carlo simulations with R for these models.In this paper,we will discussmore SDE models,including the mean reverting process,the mean reverting Ornstein-Uhlenbeck process,the square root prosess,the CIR model and Θ process.Moreover,we will make some further comments on the current study of the numerical solutionsof SDEs.

Abstract:In the big data background,not only the uncertainty of data is getting increasingly apparent,but also the incompatibility of data is becoming more and more prominent.The bipolarity as well as the fuzziness,is the instinct nature of everything.Although fuzzy set is the powerful tool to handle uncertain information,it has always neglected the incompatible bipolarity.Bipolar-valued fuzzy set introduced the incompatible bipolarity into fuzzy set theory,providing anew way to analyze and solve big data with incompatible nature,which has become a research hotspot.In this research summary,the category and characteristics of bipolarity are illustrated,then some concepts and theories are reviewed,especially the difference between bipolar-valued fuzzy sets and other corresponding fuzzy sets.Secondly,the research situations on bipolar-valued fuzzy setindomestic and world circles are discussed.Lastbut not least,the future development trend of bipolar-valued fuzzy set is analyzed.The summary provides an overall insight,as well asconstructs solid foundation for further studyinto bipolar-valued fuzzy set.

Abstract:To address the reduced face recognition accuracy in uncontrolled conditions such as the change of illumination,countenance or posture,a face recognition algorithm was proposed based on discriminative dictionary learning and regularized robust coding.Firstly,a face image is filtered by the Gabor filter to obtain the Gabor amplitude images,and the uniform local binary histogram is extracted.Then the Fisher criterion is used to gain a new discriminative dictionary,finally the regularized sparse representation is employed to test and classify the image.The experimental results based on AR face database show that the proposed algorithm has the highest face recognition rate in the existing uncontrolled environments,compared with algorithms such as Sparse Representation based Classifier,Fisher Discrimination Dictionary Learning,and Robust Sparse Coding for face recognition.

Abstract:How to evaluate the integrated economic benefits of weather service is an important and difficult issue.The definitions and calculations of direct economic benefit,relevant economic benefit,indirect economic benefit and complete economic benefit are introduced,then the evaluation models for weather service are proposed based on the input-output model,consumption factor and Leontief inverse matrix.The Shenzhen's weather service in 2012 is taken as an example to test the proposed evaluation model,which leads to some interesting and meaningful conclusions.Due to the internal relevance within industrial system,the weather service brings great relevant economic benefit,indirect economic benefit as well as complete economic benefit to economic development,which deserves more social attention and consideration.The ratio of complete economic benefit versus meteorological input ranges from 134.34:1 to 228.378:1 for Shenzhen's weather service,significantly higher than previous result of 30:1-51:1.The ratio of output versus input is relatively low in conventional industries such as agriculture,forestry,animal husbandry,fishery and building industry,yet it is high in new industries such as research & experimental development service,sci-technology exchange and promotion service,business service,resident service,telecommunication & information transmission,and computer service,which are also the key service targets for future meteorological development.

Abstract:This paper introduces the data transmission mechanism currently practiced under the cloud computing architecture.To address the lack of coupling among tasks on the links under the network transmission mechanism,we propose an improved data transmission scheme for the future network architecture which is cloud computing oriented.In view of improvement in three aspects,namely the transmission delay,transmission reliability,and storage security of the cloud node,we build a new cloud architecture in order to meet different requirements in the Qos.Being able to integrate with current network,this scheme is characterized by controllable,administrable and extensible,which makes it applicable for the new network communication infrastructure.

Abstract:A virtual desktop cloud service system deployed on VMware vSphere structure is introduced in this paper.The solution frame,network structure and user experience optimizing are detailed.Taking Binhai New Area Meteorological Service as an example,this paper emphatically described the application of virtual desktop cloud in enhancing the mobile capability of prefecture-level meteorological service,improving the resource utilization and data security,simplifying the system deployment and maintenance,as well as reducing the operating cost.This paper indicates that pieces of meteorological application software can be unified by deploying integrated service platform,which has great advantages compared with classical service management and will increase the modernization of prefecture-level meteorological service to a large extent.

Abstract:The traditional fault tree diagnosis and the script language analysis method are combined to design a circuit fault diagnosis model based on fault logic interpreter.The XML technology is employed to construct the fault tree,which is then reasoned logically by script language.An intelligent component for fault logic interpretation is created to explain the fault tree and the script language,all of which form a set of intelligent fault analysis model.After the successful software construction,the following establishment and maintenance of knowledge base and reasoning tests can be realized by electronic engineer and operator,which is clear in labor division.This power circuit fault diagnosis system is flexible and convenient in practical applications.

Abstract:Based on the knowledge of hardware structure and working principle of solar photovoltaic plant,as well as the technical parameters and the grounding performance of lightning protection equipment,this paper categorize the lightning protection grades of solar photovoltaic plant in accordance with related lightning protection standards.A method to test the lightning protection in photovoltaic system is proposed with measuring equipment of relatively large output current,and different frequency & small current test combined with help of GPS.The photovoltaic cell matrix section and power transmission & distribution equipment are chosen to be the key areas to test the lightning protection.Expected values are listed for every technical parameter and compared with measured result.This test method is low in disturbance,safe in operation,accurate in GPS locating,thus is reasonable and practical to test and judge the performance of lightning protection and grounding equipment in photovoltaic system.

Abstract:The time series of δ¹⁸O variation recorded by one reef coral from the South China Sea,two ice cores from Tibetan Plateau,and three stalagmites from central and eastern China since 1950 are analyzed and compared with meteorological data measured by local meteorological stations.The results show good consistence between δ¹⁸O records by ice core or coral and local average temperature variation,which means the δ¹⁸O records can indicate the temperature change in the past 50 years.The variation of δ¹⁸O records from stalagmites is correlated with temperature and precipitation,yet with different modes in different stalagmites.In the past 50 years,summer temperature in the Tibetan Plateau decreased during 1950s-1970s then increased from 1980s,which is indicated by δ¹⁸O variation recorded in ice cores.A continuous warming trend existed in annual sea temperature and sea surface temperature of the South China Sea,reflected by δ¹⁸O in reef coral.The δ¹⁸O from three stalagmites are all increasing in the past 50 years,the same tendency showed from δ¹⁸O in other stalagmites in Asia and the δ¹⁸O in precipitation as well.The δ¹⁸O variation in stalagmites is also in accordance with the change of summer monsoon index calculated by meteorological data,both of which indicate a weakening trend in the summer monsoon,and the trend is consistent in Asia.

Abstract:M ≥ 7 earthquakes have showed an obvious commensurability and orderliness in China's Xinjiang and its adjacent regions since 1800,with the main orderly values being(30 a)×k(k=1,2,3),11-12 a,41-43 a,18-19 a,and 5-6 a.In the guidance of the information forecasting theory of Weng Wenbo,and by combining ordered network structure analysis with complex network technology,we focus on the prediction summary of M ≥ 7 earthquakes in Xinjiang and its adjacent regions by using the ordered network structure,and add new information to further optimize the network,hence construct the 2D-and 3D-ordered network structure of M ≥ 7 earthquakes.In this paper,the network structure fully revealed the regularity of seismic activity of M ≥ 7 earthquakes in the study region during the past 210 years.Based on this,four M ≥ 7 earthquakes were predicted successfully in turn in this region,namely the Karakorum M7.1 earthquake in 1996,the M7.9 earthquake in the junction area of Russia,Mongolia,and China in 2003,and the two Yutian M 7.3 earthquakes in 2008 and 2014.At the same time,a new prediction result is presented that the future two M ≥ 7 earthquakes will probably occur around 2019-2020 and 2025-2026 in the research region.The results show that large earthquake occurred in defined region can be predicted.The method of ordered network structure analysis produces satisfactory results for the moderate and long term prediction of M ≥ 7 earthquake.