Dissertation > Excellent graduate degree dissertation topics show

Water pollution monitoring system based on cluster analysis applied research

Author: PanYuQi
Tutor: ShiBing
School: Shandong University
Course: Computer Software and Theory
Keywords: Data Mining Fuzzy Clustering Analysis Hierarchical clustering cohesion Cluster centers Cluster search Membership Matter Element Analysis Optimal Layout
CLC: TP274.4
Type: Master's thesis
Year: 2005
Downloads: 358
Quote: 3
Read: Download Dissertation

Abstract

With the rapid development of database applications and Internet applications, people demand more and more information. Faced with such a huge wealth of data resources, how to discover useful data on their own, how to obtain more important information hidden in these data, issues of particular concern. Therefore, data mining technology attendant. Cluster analysis is an important technique in data mining, clustering is to find similar data from the data set and composed of different clusters. Through clustering, one can obtain the data distribution observed characteristics of each cluster, further analysis concentrated on the particular certain clusters. This article is to study the cluster analysis method in water pollution monitoring system, discusses the four aspects including the division of rivers contaminated areas, monitoring data clustering retrieve the River Water Quality Evaluation and Monitoring Points. This paper, fuzzy clustering method is applied to the contaminated area of ??the river is divided on the first original monitoring data sample matrix, and its normalization process. Then use a new statistic - similarity calculated normalized data, the statistics contain two parameters of the distance coefficient similarity coefficient from the two aspects of the \level. Similarity matrix based clustering, each cluster represents a contaminated area. Based on fuzzy clustering, this paper studies how to retrieve monitoring data clustering. Clustering retrieval by clustering similar data classification, data query and on this basis, so as to improve the efficiency of data retrieval. The clustering search process: first data clustering, clustering results calculated the center of each cluster, a cluster center file, and then on this basis, by comparing the cluster center with queries data for data retrieval. Application of fuzzy evaluation method to evaluate the river's water environment quality, mainly is through the establishment of a reasonable membership function, according to the actual concentration value to determine various pollution indicators of the right weight, to ensure that the evaluation of the results of the objective, and accurate, the pollution indicators of the right to re-set with membership degree matrix multiplication, each monitoring site evaluation matrix, monitoring points upstream river length is controlled by the total length ratio multiplied by the evaluation matrix, we can draw the full set of comprehensive evaluation of the river. By matter element analysis method and fuzzy clustering analysis method, this paper presents a fuzzy clustering analysis method based on matter-element analysis of the correlation function. The essence of the method is based on the matter element analysis resulting correlation function as the similarity between the sample statistic, which construct fuzzy relationship matrix, in order to achieve the sample fuzzy clustering analysis. This paper the method is applied to water pollution monitoring system to optimize the layout of experimental data proved to be feasible. Finally, the article discusses the inadequacies of the existing water pollution monitoring system in the cluster analysis applications and improved.

Related Dissertations

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Automation systems > Data processing, data processing system > Centralized testing and roving detection system
© 2012 www.DissertationTopic.Net Mobile