A method of inserting ActiveX controls indirectly to the customized file dialog is introduced.It solves the problem that ActiveX can not be inserted directly to the common file dialog of MFC,which is discommodious for applications development.The example presented has worthful reference value and application value.

Three common evaluation functions for automatic focusing were introduced. They were deviations function, gradient function and frequency-spectra function, the later two were optimized. Their merits and deficits in delicacy, stability and velocity were analyzed. And a multiple automatic focusing algorithm based on three-stages was proposed herein. It performs well in the video measuring system and can focus exactly and rapidly.

This paper clarifies three methods of Internet connection of the existing embedded systems. After analysis and comparisons of their advantages and disadvantages, a new method for virtual TCP/IP peripheral-based Internet connection of intelligent terminals is discussed and applied to our fingerprint identification module. Experimental result shows that this method is easy and feasible to design application software.

研究了科技项目招标投标过程中的标书评价问题 ,给出了一种基于灰色趋势关联分析的评标决策模型 ,并且以某新产品计划科技招标项目的评标工作为例 ,基于评标决策模型用实际数据对新产品计划科技招标项目进行了评标研究 .

Grey system theory is an effective tool for system analysis, modeling, prediction decision making, and control on poor-information systems. This paper discusses the application of grey system theory to data mining in poor-information systems in which there is a little or non-complete samples, or sample data updates quickly, or whole sample data is very complex, but in some space or time region, sample data obeyes regularity. Some new algorithms, i.e. grey relational degree algorithm, grey statistics algorithm, grey clustering algorithm, and grey statistic clustering algorithm have been presented. And the system structure of grey data mining systems is constructed.

阐述了现有各种嵌入式系统的互联网接入方案,分析和比较了各种接入方式的优缺点后,重点阐述了一种基于TCP/IP的智能终端接入技术方案.

The characteristics of poorinformation in database and the requirement of poor information data mining are analyzed. Based on gray trend relational degree and general system theory, the model of gray trend relational system is investigated. As an application of this model, the gray trend clustering method is presented.Based on scanty data of history sale of a large wine Co. Ltd, according to increasing tendency of sale goods, all kinds of goods are clustered by gray trend clustering method, and the clustering result accords with sale fact. It shows that the gray trend clustering method is valid, convenient and practical.

利用人工神经网络的方法实现系统云灰色模型的参数白化,提出了系统云灰色神经网络模型SCGNNM(1,1),并给出了相应的学习算法.仿真结果证明了模型的可行性.

将AHP方法运用到层次多目标群决策中,确定多层目标准则体系中的目标层和准则层的权系数,符合判断指标一致性的需要,而对各方案关于各个准则的权重,采用模糊语言进行评价,满足决策实用性的需求.结合模糊最大最小集法和上述方法,提出了一种带有偏好的模糊多层多目标群决策方法,并用算例对其求解过程进行了说明.

根据人工神经网络的结构特点和信息传输与处理特征,提出了利用物理神经网络作为通信信道的观点.在分析了现有通信信道物理模型之后,具体讨论了作为通信信道的神经网络的可能拓朴结构、神经网络学习要点、允许传输信号的模式特征,研究了神经网络信道的传输延时、通信带宽及信噪比,分析了神经网络信道处理多媒体信号的计算潜力,并着重剖析了神经网络信道固有的保密通信特征.最后给出了一个具体的神经网络信道设计实例.

结合灰色趋势关联度,利用数学一般系统理论,提出了一般灰色趋势关联系统模型,作为该模型的应用,对灰色趋势关联分类、灰色趋势关联聚类、灰色趋势关联预测等三种应用方法进行了研究.

以数据仓库为基础的在线联机分析处理(OLAP)技术是决策支持系统(DSS)中一种新的决策分析方法.该文从DSS的实际需求出发,分析了数据仓库的特征和结构模型,研究了OLAP的特征和体系结构及其实现决策分析的过程,最后给出了应用实例,并探讨了完善DSS结构的研究方向.

从决策支持系统的一般性出发,利用一般系统理论中的MT理论对决策支持系统的一般功能要求进行了分析,提出了一般决策支持系统的架构体系,并在该架构的指导下进行了具体问题的分析,为DSS的研究与发展引进了一种新思路.

In this paper, some models of conventional information retrieval are described firstly; also algorithms of those models are included and evaluated. Hierarchy structure of information retrieval is summarized subsequently. Lastly, research advances, key technologies and prospects are introduced, and some valuable problems are brought forward for further study.

Based on fuzzy evaluation for every sub-objective importance in a multi-objective optimization by each expert, combining the preference of a decision maker and the genetic algorithm, a multi-objective optimization algorithm with fuzzy preference is proposed. This algorithm utilizes the fuzzy maximizing set and minimizing set to turn a real number value into a fuzzy number, and to obtain the weight of every sub-objective in multi-objective optimization problem. Then a multi-objective optimization problem is transformed into a single objective optimization problem, and an optimization solution is gained by using the genetic algorithm. Finally, a study case is given.

分析了决策支持系统(DSS)的结构特点,针对传统DSS的不足,研究了数据仓库的结构特征,进而提出了一种基于数据仓库的融合式决策支持系统(HDSS)的结构,并从系统管理和系统功能两个角度对HDSS的结构作了详细介绍.

The development of decision-making methods and characteristics of the traditional DSS(Decision Support System) are described. The architecture of a Hybrid DSS based on dataware is offered. This architecture is detailed from two aspects: management structure and function structure of a system. The operational structure and process of this HDSS is also introduced.