【作者】 宋记锋；

【导师】 葛运建；

【作者基本信息】 中国科学技术大学 ， 模式识别与智能系统， 2008， 博士

【摘要】 信息链由信息获取、信息传输和信息传处理三个环节组成,他们构成了信息获取科学的三大学科支柱,即信息获取、计算机、通讯。目前,信息处理和信息传输方面的学科建设同趋成熟和完善,而信息链的源头—信息获取仍旧停留在传统的传感技术和检测技术等较低层次,逐渐成为21世纪信息科学与技术发展的瓶颈。创建“信息获取科学与技术学科”,是信息化时代在传统的传感技术和检测技术基础上进行学科升华,它在科学的深度和技术的广度上突破了原来的界限,是一个具有基础性、前瞻性、战略性的创新研究方向,有重要的意义。已经有不少学者对此进行了研究,状得了相应的成果。本文力图沿着这个创新方向,在前人的基础上,研究信息获取科学的若干基本问题,希望能够为以后的学科建设起到抛砖引玉的作用。本文主要做了以下几个方面的工作:一,以溯源论的思想,对传感技术和检测技术的物质层进行了内在基本原理分析。从微观粒子运动和物理基本相互作用的角度,针对物质层进行信息获取概念的界定,指出信息即物质存在形式的外在表达,信息获取即通过一系列手段对信息进行转换和捕获。这种概念源自物质的基本存在形式,适合于传感器领域的普适性分析。二,在物质层研究中,从微观粒子运动角度切入,对常见敏感效应进行了剖析,基于物质粒子运动形式和信号传递机制的区别,提出一种传感器二元法分类,提出了传感器族属概念:并在此基础上挖掘出相应的敏感机理族群效应。该族群效应具有几个明显的特征:1,不同族传感器性能差异明显;2,同族传感器其性能相近;3,传感器族属关系越近则性能差别越小。针对传感器物质层进行了信号传递分析,建立了一个多环节误差传递数学模型,并基于此给出了相应的误差累计传递函数及相应的性质,用于评估误差在物质层的累计程度,给出一种通过减少环节来降低误差的设计策略。三,在信息获取数据层研究上,针对数宁滤波用的差分方程,给出了一种完全线性化的变形Z变换降阶算法。该算法本质上是通过对Z变换和逆变换进行线性化处理,避免求导和留数运算,仅涉及行列式和矩阵算式,非常适合计算机自动求解,并且结果为具有重要分析意义的封闭解。该算法针对不同的方程特征根,依次对输入项进行拆分处理,利用递推操作对差分方程进行降阶表达,再结合矩阵和行列式工具运算,直至获得最终封闭解,并给出了建立在克莱姆法则上的封闭解判据。该算法的严格证明和实例分析都一并在文中得到详细阐述。四,在信息获取特征层研究上,提出一种利用模糊数学实现瓦斯突出特征快速识别的算法。并以国家自然基金重点项目(煤矿瓦斯传感技术和预警信息系统基础理论与关键技术研究No:50534050)为应用背景,分析了煤矿瓦斯突出与正常涌出的浓度数据,通过深度挖掘瓦斯浓度变化的特征,建立瓦斯突出浓度模型,利用模糊推理区分瓦斯涌出与瓦斯突出,实现瓦斯突出预警;瓦斯突出预警装置模拟试验显示,瓦斯突出能够得到有效预警。该系统能够比较方便的与煤矿瓦斯监测系统直接接驳,对煤矿安全生产有一定的应用价值。更多还原

【Abstract】 The information-chain is made up of information acquisition, processing and transportation, which form the three mainstays of Information Science, namely. Information acquisition, Computer and Communication. At the present time, discipline systems of information processing and information transportation have already been constructed. However, the source of information chain, information acquisition still rests on the low level of sensor technique, which is becoming the bottleneck of Information Science in the 21^st Century.To construct the discipline system of information acquisition, is to upgrade the traditional sensor technique to the level of science, and is to break through the traditional limitation in scientific depth and technical scope. This is a new scientific research domain of leading importance. Therefore, information acquisition is gaining in popularity as evidenced by the increasing number of researches. This thesis tries to research on some questions in Information Acquisition Science in this new domain. We wish an introductory work for Information Acquisition Science.The main works we do in this thesis may be summed as follows:1, From microscopic viewpoint, the origin of sensor technologies is the basic interacting effects between particles. The analysis of means about material layer of information acquisition has carried out. On the groundwork of particle-field interaction, the concept of information of material layer is defined as "external expression of existing form of matter", and the concept of information acquisition is defined as the process of information convert and capture. Due to its basically physical background, this concept is suit for general analysis for sensor technologies.2, This research proposed a symbolic expression method which classifies sensory mechanisms by Boolean symbolism. This classification belongs to the universal model research of the material layer of the discipline of information acquisition science and technology. Based on basic physics theories, sensory mechanisms were classified using a binary tree structure. Every type mechanism has been expressed with Boolean symbol. The symbolic expression method describes the performances deference of sensors, including accuracy and dynamic performance. Some conclusions could be easily obtained from this group effect: 1, different group has different performances in information acquisition; 2, same group has similar performances; 3, more closer of relationship between groups, more similar on performance; 4, size has an infection on frequency performance of sensors. Actual applications were analyzed to demonstrate the relation concluded from the Boolea（?） classification presented in this thesis. The results showed that this symbolic expression method could be used to forecast the performances of new sensor on a certain extend, especially in inertial sensor field. The process of signal transport through the material layer is analyzed, and a poly layer model of error transmission is established and adopted to interpret the precision limitation of sensor technologies. A design strategy for high-accuracy survey is proposed based on this poly layer model of error transmission.3, It is difficult to get the analytic solution of difference equation by using traditional methods with computer. A linear method based on complete linear transformation for getting the analytic solution of difference equation is proposed in this thesis. The idea of this new method is dimensionality reduction. The orders of the high-order difference equations are reduced through determinant calculation, and this process is repeated till the last analytic expression of the solution of the difference equation is gotten. In essence, this algorithm is a transformation of Z-transform method combined with linear algebra method. The linear property of this algorithm enables one to get analytic solution other than numerical solution of difference equations using computer. All corresponding theorems and characters are proved and illustrated.4, Supported by the National Nature Science Foundation of China （Project 50534050）, a real-time alarm method for methane outbursts is proposed. The method is established on the real-time analysis of the increasing velocity and the acceleration of the methane concentration curve. A real-time alarm algorithm based on fuzzy mathematics is presented. A prototype instrument and experiment results are presented and analyzed, which show good performance of this proposed method. This method can be used to give triggering signal to the explosion suppressing system at very early stage of outburst. Considering the actual status of china, this device and the corresponding method is suit for linking with the methane monitoring system of mine, which makes this research valuable for safety in production.更多还原