Abstract:For multivariate pseudo-linear regressive moving average systems,a multivariate extended stochastic gradient(ESG) algorithm is discussed.In order to reduce the computational cost of the identification algorithm,we decompose a multivariate system into several subsystems,and derive a partially coupled(subsystem) ESG algorithm and a partially coupled(subsystem) multi-innovation ESG algorithm according to the coupling identification concept and the multi-innovation identification theory.Furthermore,we extend these methods to multivariate pseudo-linear autoregressive moving average systems and present a partially coupled(subsystem) generalized extended stochastic gradient(GESG) algorithm and a partially coupled(subsystem) multi-innovation GESG algorithm.The computational efficiencies of the multivariate ESG algorithm,the partially coupled ESG algorithm and the partially coupled multi-innovation ESG algorithm are analyzed.

Abstract:The force telepresence telerobot system is a typical human-robot-environment-interaction system.The environment as a terminal unit of the whole teleoperation system is the object perceived and manipulated by the telerobot.The environment dynamics directly affects the control performance of the teleoperation system.This paper analyzes the characteristics of environment dynamics,then proposes three kinds of dynamics models of environment under three contact conditions such as collision contact,linear steady contact,and non-linear steady contact,and gives their equivalent impedance forms,respectively.The research results in this paper provide a basis for the force telepresence telerobot system analyzing.

Abstract:The three-dimensional synchronous meteorological data are demanded for further understanding the occurrence,development and disappearance mechanism of meteorological phenomena on the atmospheric boundary layer.However,current detection technology seems powerless to this requirement.A new detection mode is proposed in this paper,which is applying the multi-rotor Unmanned Aerial Vehicles(UAVs) to autonomously detect the meteorological elements in the atmospheric boundary layer.Furthermore,its background,significance,characteristics,advantages,procedure,feasibility and some scientific problems related to this detection mode are analyzed.This detection mode has the characteristics of high pertinency,flexibility,mobility and maneuverability,and the functions of autonomous dynamic synchronous detection,intelligent Auto-Tracking detection.

Abstract:Through the real-time analysis of Doppler radar PUP data,including Echo Tops (ET),Radar Reflectivity (R19),and Vertically Integrated Liquid (VIL) water content,ArcGIS Identify tool is programmed,which will get data collection from raster feature generated by radar data.We propose the concepts of warning radius and warning region attached angle θ,design monitoring point sampling algorithm,azimuth algorithm,(θ,α) point selection algorithm and identify algorithm,furthermore,complete the simulations of each algorithm based on C#.Finally,the automatic warning system prototype was accomplished in Visual Studio 2008,which can generate real-time operating parameters of each operating point.The research provides necessary theoretical and technical support for intelligent operation of weather modification.

Abstract:Based on Chaozhou meteorological data including sunshine hours,total cloud amount,precipitation,surface vapor pressure,mean temperature,mean wind velocity and relative humidity from meteorological stations during 1969—2011,the characteristics and influencing factors of evaporation change in Chaozhou were studied with the methods of evaporation anomaly percentage,accumulative anomalies,Mann-kendall test,and wavelet analysis.The result indicates that the evaporation of 1970s and 1980s in Chaozhou were the most for the period of 1969—2011,while the evaporation of 1990s was the least.Seasonal fluctuation was obvious in evaporation change.Mutation test shows that the sunshine hours increased rapidly in 1970s,yet without obvious mutation.According to the period analysis of annual and seasonal evaporation changes,the evaporation now is relatively high in the more significant time scale.The decrease of annual evaporation is closely related with the increase of mean temperature,and may also be related with the increase of sunshine hours in summer.

Abstract:According to the risk management theory and method,this paper identifies and analyzes the meteorological disaster risks of tobacco plantations in Longyan,Fujian province.On the basis of that,by analyzing the factors of hazard,exposure and vulnerability,a risk index system and a risk evaluation model are constructed.By using the risk index system and the risk evaluation model for each county in Longyan,the frost and flood disaster risk indexes can be obtained for each county.According to the risk classification standards,the levels of risk zoning can be provided and used as references for structure adjustments in tobacco planting and disaster prevention in Longyan.

Abstract:Typhoon is meso-α scale system and sometimes comparable to large scale system.The isolation of typhoon from the environment flow is an important step in order to investigate the interaction between the environment flow and typhoon circulation.Three methods including time filter,space filter and dynamic method are applied to separate the typhoon Megi (1013) of year 2010 from the environmental flow.The separation results show that,time filter and space filter methods have relative randomness and are somewhat subject to the life cycle and the size of typhoon,which will leave residual typhoon wind to the environmental field,thus weaken the environmental flow;while the dynamic method can well separate the typhoon from the environmental flow.Both advantages and disadvantages exist in the three methods.The time filtered wind field is well balanced with other time filtered fields such as temperature and height.While when using space filter,wind may not be consistent with height due to the different scale between wind,height and temperature field.Nevertheless space filter is relatively simple to use and only need one time series,while time filter needs longer time series,which result in no application of time filter in operational forecasting.Dynamic method is better in performance,but relatively complex and sometimes prone to imbalance in lower level and has much space to improve.

Abstract:Data of lightning location (2007—2010) and thunderstorm day (1961—2010) from 52 ADTD lightning detection stations are used to analyze characteristics of lightning parameters for five provinces (Hunan,Hubei,Anhui,Jiangxi,Henan) in central China.Results show that lightning parameters present different variation patterns in latitudes on the northern and southern side of 30.8±0.2°N.In latitude from 36.4°N to 30.8°N,the positive lightning intensity descends,with small positive lightning gradient change and average value of 16.48 kA·μs^-1.While the negative lightning intensity and gradient reach maximum at 32°N and then fall rapidly at 30.8°N.The occurrence time of maximum and minimum lightning frequency shows a decentralized appearance in a day,and the thunderstorm day starts from approximately April 5 and ends at approximately September 17 in a year.While in latitude from 30.8°N to 24.4°N,both the positive and negative lightning intensity increase,but the positive one increases rapidly from 52.11 kA to 77.88kA and the negative one increases slowly.The positive lightning gradient has small change with average value of 14.24 kA·μs^-1,and the negative lightning gradient decreases from 12.07 kA·μs^-1 to 8.90 kA·μs^-1.Most lightning occurs on 15:00—16:00,and the least lightning occurs on 8:00—10:00 in a day,the thunderstorm day starts from about February 18 and ends at about October 17 in a year.

Abstract:Based on the lightning data of Anhui province from 2005 to 2010,taking the administrative region (city) as the basic unit,the risk of lightning disaster in Anhui province was analyzed.Thunderstorm day (T),Lightning density (N),population density (L) and GDP per capita (D) were introduced as main assessment indices and Analytic Hierarchy Process (AHP) model was adopted to measure the weight.The lightning risk zoning of Anhui province was obtained by the model,and previous lightning disasters were used to verify the validity of the zoning.The result shows that cities of Huangshan,Ma'anshan,Tongling,Chizhou,Anqing,Huainan,Hefei,Fuyang belong to the high risk areas of lightning disaster,while cities of Suzhou,Bozhou,Huaibei belong to the low risk areas of lightning,and other cities belong to the middle risk area of lightning disaster.

Abstract:Cangnan county of Zhejiang province is chosen as a sample area to establish an effective method of lightning disaster risk zoning.Monitoring data of lightning occurred in Cangnan during 2007—2011,statistics of lightning disasters and related basic information during 1996—2011 are used to analyze the characteristics of lightning disasters in Cangnan by using the method of Analytic Hierarchy Process-Grey Relational Analysis(AHP-GRA).The AHP is used to determine the weight of evaluation index,while the GRA is used to calculate the relational coefficient between different evaluation indexes.The comprehensive weighted relational coefficients are then used to derive a reliable lightning disaster risk zoning.The results show that the risk grades in descending order are very high risk zone,including Lingxi town and Longgang town; high risk zone,including Yishan town,Qianku town and Jinxiang town; medium risk zone,including Fanshan town,Mazhan town,Zaoxi town,Qiaodun town; and low risk zone,including Chixi town,Dailing town,Fengyang town.The risk zoning results in Cangnan county are verified through ArcGIS information of actual lightning disasters.The verification result shows that the AHP-GRA method is effective for lightning disaster risk zoning.This research will provide reference for the planning of lightning disaster prevention and evaluation of lightning disaster risk for major projects.

Abstract:According to the latest statistical data of hydrology,a total of 21 floods took place in the Changjiang River Basin from 1827 to now,which showed a very obvious orderliness.Based on the information forecasting theory and ordered network analysis,we construct the 2D-and 3D-ordered network structure and make prediction research.Prediction results show that the future big deluges will probably occur in the Changjiang River Basin around 2014,2020,2030,2036,2051 and 2058.

Abstract:In this paper,a generalized stochastic model of high-order Hopfield-typeneural networks with time-varying delays and distributed parametersis considered.The sufficientconditions ensuring the exponential stability of the systems are developed by using Lyapunovstability theory,an integral inequality and Halanay's inequality.The proposed conditions are diffusion-dependent due to the use of the new integral inequality.As a result,the obtained conditions may have some advantages over the those previously reported.As anillustration,an numerical example is worked out using the resultsobtained.

Abstract:This paper studies an agent's consumption/portfolio and retirement problem,in which the dividend-payment of risk assets as well as the utility loss from labor are considered.The agent has an option to retire from his work.Before retirement the agent receives labor income but suffers a utility loss due to work,however,by deciding to retire from work,he saves the utility loss but gives up labor income.The agent utility comes from consumption,which is directly influenced by utility loss due to labor.We obtain an explicit solution for the agent's critical wealth level and optimal consumption/portfolio strategy by using dynamic programming methods to solve a free boundary value problem.