不完备多源信息融合的粒计算方法研究

发布时间：2018-05-11 18:27

本文选题：不完备 + 条件熵　；参考：《重庆理工大学》2017年硕士论文

【摘要】：由于人类社会的不断发展与进步,人们获取数据的方式越来越多样化,面对形式多样的、数量巨大的、关系复杂的、要求及时处理的这些数据,如何得到有用的删除冗余的信息,如何将得到的信息提炼后得到更精确的信息,是当前的研究热点之一,在电子科技和网络工程的很多领域中,得到精确和完备的信息是很困难的,采集到的数据通常包含噪声,模糊且不完整。那么,当采集到的信息都是模糊或不完整的情况下,如何将不完备的多个信息源进行融合就成为了多传感器信息融合技术的重点。随着科学研究的不断深入,对不同环境下各种数据的处理就显得尤为重要,本文基于此背景,在不同环境下的序信息系统中,通过逻辑运算算子构建了几种新的粗糙模糊集模型,并研究了多个模糊或不完整的信息源如何进行融合的方法。主要创新点如下:1.在序信息系统以及直觉模糊序信息系统中,将变精度和程度粗糙集模型通过逻辑运算算子结合起来构建三种新的粗糙模糊集模型,并通过实际案例验证了所构建模型的有效性。2.研究了当多个信源都是不完备时,也即是当每一个信息系统均为不完备的信息系统时,如何进行信息融合的方法,并根据提出的融合方法设计了对应的算法。然后,基于UCI数据集设计了一系列实验,将本文提出的融合方法和传统的融合方法进行比较,进一步验证了本文提出的融合方法在精度方面具有十分明显的优势。3.将不完备信息系统的融合方法应用到模糊信息系统中,通过定义一种新的相似性度量,构造了一种新的相似二元关系,进而在此基础上建立了融合模型。
[Abstract]:As a result of the continuous development and progress of human society, the ways in which people obtain data are becoming more and more diversified. In the face of these data, which are diverse in form, large in quantity, complicated in relation, and require timely processing, How to get useful redundant information, how to extract the information and get more accurate information is one of the current research hotspots, in many fields of electronic science and technology and network engineering, It is difficult to obtain accurate and complete information. The collected data usually contain noise, blur and incomplete. Then, when the information collected is fuzzy or incomplete, how to fuse incomplete multiple information sources has become the focus of multi-sensor information fusion technology. With the development of scientific research, it is very important to deal with all kinds of data in different environments. Several new rough fuzzy set models are constructed by logical operators, and how to fuse multiple fuzzy or incomplete information sources is studied. The main innovations are as follows: 1. In order information system and intuitionistic fuzzy order information system, three new rough fuzzy set models are constructed by combining variable precision and degree rough set models through logical operators. This paper studies the method of information fusion when many information sources are incomplete, that is, every information system is incomplete, and the corresponding algorithm is designed according to the proposed fusion method. Then, a series of experiments based on UCI data set are designed, comparing the fusion method proposed in this paper with the traditional fusion method, and further validating that the fusion method proposed in this paper has a very obvious advantage in accuracy. The fusion method of incomplete information system is applied to fuzzy information system. By defining a new similarity measure, a new similarity binary relation is constructed, and then a fusion model is established.
【学位授予单位】：重庆理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP202

【参考文献】