中国概率统计学会(CSPS)和Institute of Mathematical Statistics(IMS)2005年举办CSPS/IMS联合会议征文通知

经CSPS与IMS商定将于2005年在北京联合召开概率统计学术会议.该会议由中国概率统计学会和IMS联合主办,中国现场研究会、中国统计学会和中国统计教育学会协办.现已决定于2005年7月9日-12日(8日报到)在北京大学英杰交流中心举行这次“CSPS/IMS 2005联合会议”.有关会议事宜如下: 1.投稿论文须是2005年7月底以前没有发表的.

Mathematical Statistics Method of Residual Static Torque Controlling

This paper analyzes the problems existing in the detection and determination on the traditional residual torque. It is pointed out that the detection value is mistakenly used as the truth value and the lower limit of design torque is mistakenly used as the residual torque criterion. The calculation formula of the determination （limit） value of the test value of RM and DFM is proposed. By defining related parameters,the torque limitvalue corresponding to the tightening level, the ratio of the measured value to the truth value, μ and σ related tothe ratio of A and C, and the confidence coefficient related to the confidence probability, a mathematical statistical method of residual static torque controlling is established, which improves the control precision on the residual torque.

Potential hydraulic connectivity of coal mine aquifers based on statistical analysis of hydrogeochemistry

Mining activities interfere with the natural groundwater chemical environment,which may lead to hydrogeochemical changes of aquifers and mine water inrush disasters.This study analyzed the hydrochemical compositions of 80 water samples in three aquifers and developed a water source identification model to explore the control factors and potential hydraulic connection of groundwater chemistry in a coal mine.The results showed that the hydrochemical types of the three aquifers were different.The main hydrochemical compositions of the loose-layer,coal-bearing,and limestone aquifers were HCO_(3)·Cl-Na,SO_(4)·HCO_(3)-Na,and SO_(4)-Na·Ca,respectively.The correlation,Unmix,and factor an-alyses showed that the hydrochemical composition of groundwater was controlled by the dissolution of soluble minerals(such as calcite,dolomite,gypsum,and halite)and the weathering of silicate minerals.The factor score plot combined with Q-mode cluster analysis demon-strated no remarkable hydraulic connection among the three aquifers in the study area.The water source identification model effectively identified the source of inrush water.Moreover,the mixing ratio model rationally quantified the contributions of the three aquifers to inrush water.

Rapid Determination of Complete Distribution of Pore and Throat in Tight Oil Sandstone of Triassic Yanchang Formation in Ordos Basin, China

This study aimed to investigate the complete distribution of reservoir space in tight oil sandstone combining casting slices, field emission scanning electron microscopy(FE-SEM), the pore-throat theory model, high-resolution image processing, mathematical statistics, and other technical means. Results of reservoir samples from the Xin’anbian area of Ordos Basin showed that the total pore radius curve of the tight oil sandstone reservoir exhibited a multi-peak distribution, and the peaks appeared to be more focused on the ends of the range. This proved that pores with a radius of 1–50,000 nm provided the most significant storage space for tight oil, indicating that special attention should be paid to this range of the pore size distribution. Meanwhile, the complete throat radius curve of the tight oil sandstone reservoir exhibited a multipeak distribution. However, the peak values were distributed throughout the scales. This confirmed that the throat radius in the tight oil sandstone reservoir was not only in the range of hundreds of nanometers but was also widely distributed in the scale approximately equal to the pore size. The new rapid determination method could provide a precise theoretical basis for the comprehensive evaluation, exploration, and development of a tight oil sandstone reservoir.

NON-LINEAR STOCHASTIC RESPONSE AND ITS EXTREME VALUE OF JACK UP PLATFORMS

The non-linear stochastic response of a jack-up platform subjected to wave load has been analyzed dynamically in this paper, and the analysis method in time domain is considered. Monte Carlo simulation is used to generate random sea. An emphasis is placed on the nonlinear hydrodynamic force. Several distributions for the statistical estimation of extreme responses are compared. For Gumble distribution, the parameters of its asymptotic distribution expression have been checked. The results show that the Gumble distribution agrees well with the simulated values of the responses.

Study on Lubricating Oil Monitoring Technology

Lubricating oil monitoring has been proven to be an effective method for detecting and diagnosing machinery failures and essential for realizing condition based maintenance. In this paper, mathematical statistics methods for determining the oil parameters featuring machinery failures and the parameters＇ probability distribution functions and their thresholds are put forward.

A Support Vector Machine Based on Bayesian Criterion

SVM handles classification problem only considering samples themselves and the classification effect depends on the characteristics of the training samples but not the current information of classified problem.From the phenomena of data crossing in systems,this paper improves the classification effect of SVM by adding the prior probability item reflecting the classified problem information into the decision function,which fuses the Bayesian criterion into SVM.The detailed deducing process and realizing steps of the algorithm are put forward.It is verified through two instances.The results showthat the new algorithm has better effect than the traditional SVM algorithm,and its robustness and sensitivity are all improved.

Application of Digital Analysis Technology in Refractory Petrographic Analysis

Six specimens of fused cast AZS 33#material were taken as samples,and their chemical compositions and physical properties（ initial separating temperature o glass phase,corrosion amount in 24 h,glass phase seepage,and bulk density） were tested; low powered backscattered electron images of the samples（ three for each sample） were taken by a scanning electronic microscope with a reinforced backscattering probe. Then,firstly,the images were processed,and the phase compositions（relative contents of pores,matrix,alumina- zirconia eutectoid and primary baddeleyite） of the samples were obtained by counting. Secondly,the correlation coefficients of all parameters were calculated; indistinctive coefficients and unimportant variables were eliminated by F- test. Thirdly,the roadmap of the parameters was established and explained according to professiona knowledge.

Experimental Research on Influence of Some Factors on Ejection Height of Cargo Projectile

Some factors influencing the cargo projectile's ejection height in different conditions were analyzed by using harmonic curve diagram and cluster method.The test results of the ejection height for a certain cargo projectile show that the wind speed and direction are the main influence factors.For given meteorological conditions,the applied environmental stress becomes the main cause,while the launching angle does not influence the ejection height.

Spatial Distribution of Hospitals in Handan City and Its Influencing Factors

Hospital is an important factor of people’s livelihood security,and the spatial layout of hospitals effectively ensures the medical convenience for residents.Location entropy and mathematical statistical analysis are used to study spatial distribution of hospitals.The results display that the distribution of medical facilities in Handan City is at a disadvantage level in Hebei Province,and medical facilities arr concentrated in the plain area.The layout of grade 3A hospitals in Hebei Province is characterized by urban centralization,and it is stronger in the east and weaker in the west.There is no medical facilities in Feixiang District of Handan City,and layout of medical facilities in Hanshan District and Congtai District is at advantage level of Handan City.The built-up area is the influencing factor for the distribution of medical resources.

Application of Multi-Attribute Crossplots to Tight Gas in X Area, Western China

The gas-bearing reservoir in X area is mainly the tight sandstone reservoir characterized by low porosity and permeability, frequently lateral variation and poor connectivity of single sand. The previous research results reveal that the general seismic attributes analysis cannot meet the requirement of fluid identification. This is because the relationship between seismic attributes and their implication is uncertain and ambiguous, which decreases the precision of both reservoir prediction and fluid identification. To overcome the problem, multi-attribute crossplot technology is proposed from the mathematical statistical point of view rather than the correspondence between the seismic attributes and their geological implication. In this method, the wells which have the same statistical law are classified firstly, and then all the interest wells are retained while the wells beyond the statistical law are eliminated, and the seismic attributes sensitive to the same types of eliminated wells are optimized and used to generate crossplots. The nonzero area of their crossplots results just predicts the potential distribution. The discontinuity of subsurface geological conditions results in the non-continuous shape and the seismic bin lead to the mosaic form. The optimization of sensitive attributes relative to the same types of wells is independent from each other, and thus the order of attributes in crossplots does not affect the final prediction results. This method is based on the statistical theory and suitable for the areas such as the study area abundant of lots of well data. Application to X area proves the effectiveness of this method and predicts plane distribution about different types of gas production. Due to the effect of faults and other geological factors, the partition prediction results using multi-attribute crossplots reach 95% of coincidence which is obviously and far higher than the results of the whole area. The final prediction results show that the potential areas with medium and high gas production are mainly concentrated in the northern part of the study area, where lots of development research will be strengthened.

Why SA Can Beat the Exponential Explosion in Heuristic Search

In tree (or graph) search,most algorithms mainly use the local heuristic information of each indi- vidual node.But in the statistical heuristic search algorithms the global information about subtrees is used effectively so that the computational complexity is greatly reduced.In this paper the problem of how the global information can be extracted from the local one is discussed.Some features of SA are also concerned.

The Distribution of Multinomial Logit Model with Combined Errors of Perception and Measure

Logit\|Probit model is investigated as an extra measured error mixed with the user′s utility in multinomial logit model. In the situation, the discrete choice proportion is considered as the conditional probability, the form of multinomial logit and the random variable. The distribution of the conditional probability is derived from the multinormal distribution of the measured error to the distribution of quotient of lognormal random variables. Examples are employed to illustrate its application.