Dissertation > Excellent graduate degree dissertation topics show

Study on Classification Association Rules Mining and Its Application in Complicated Industry Process

Author: RenJia
Tutor: ZuoJianï¼›SuHongYe
School: Zhejiang University
Course: Control Science and Engineering
Keywords: Complex industrial processes Data Mining The multivariate monitoring type discretization Fuzzy classification association rules Fuzzy Systems Fuzzy path query system Adsorption Separation Process WSA gas sulfuric acid process
CLC: TP311.13
Type: PhD thesis
Year: 2006
Downloads: 689
Quote: 7
Read: Download Dissertation

Abstract


Current automation and database technology has penetrated to a wide range of industrial processes, a large number of systems running data accumulated in these processes, real-time database, which contains many of the industrial process control, parameter optimization, product quality and production management information they provide a broad platform for the application of data mining technology in the field of complex industrial processes. In this thesis research background of complex industrial processes, classification association rule mining algorithm for the study of the main line, contains several important stages in the process of data mining technology and content (data preprocessing technology, exploratory analysis methods, data mining modeling and implementation of applications). Combined with the actual project (a lead smelting enterprises sintering flue gas WSA acid production process mining association rules), the specific implementation and application. The research work of this paper is summarized as follows data preprocessing stage of a multi-variable supervision type discretization (MSD) algorithm for the characteristics of the industrial process data (multi-variable, strong coupling, large amount of data processing), the paper propose a multivariable supervision type discretization algorithm. The algorithm is divided into two levels: First, the use of clustering algorithm for coarse discretization (fully taken into account in the discretization process the dataset multiple condition variables related information); then use Chi2 discretization algorithm fine discretization (the discrete process fully absorb the data set of classified information). The discretization algorithm takes advantage of the distribution of the data set characteristics and classification of information, discretized data sets automatically. Comparison test analysis and practical application show that the algorithm has good discretization. In the modeling stage, three kinds of data mining model (1) mining (FCARM) model in the existing classification association rule mining algorithm based on a fuzzy classification association rules, this paper presents a fuzzy classification association rule mining model. The model contribution as follows: First, the introduction of a new definition of distance-based support; Second, the multivariate supervision discretization algorithm is introduced to the discretization process of continuous attributes discretization effects enhanced model. Third, to further the concept of fuzzy set introduced into the model, to overcome the disadvantage of excellent boundary of the property division process. Finally, the use of this model to the aromatics extraction industrial process history data mining. (2) proposed a fuzzy systems based on fuzzy classification association rule set (fuzzyCARs) construction method, there are two more difficult to overcome the difficulties in traditional fuzzy system modeling process: One is a data object

Related Dissertations

  1. Study on Fault-tolerant Guaranteed Cost Control for Nonlinear Systems,TP13
  2. Stability Analysis and Controller Design for 2D Fuzzy Systems,TP13
  3. Research on Water-saving Suitable Technique Model Based on Fuzzy Neural Network and System Development,S274.1
  4. Neutral fuzzy systems control non- friable,TP13
  5. Research on Fuzzy Control of Nonlinear Systems,O231
  6. The Modeling and Application of Niche and Community Based on Type-2 Fuzzy Sets,Q141
  7. Research on Fuzzy System and Neural Network and Its Applications in the Thermal Process,TK123
  8. Research on Control Approach of Fuzzy Discrete Systems with Time Delays,TP13
  9. Delay-dependent Stability of TS fuzzy system,TP13
  10. Data Collection and Assisting Force Analysis of Sit-to-Stand Transfer Assisting Robot,TP242
  11. An Extended Cultural Algorithm,TP18
  12. Robust control for a class of uncertain discrete generalized fuzzy systems,TP13
  13. Based on Dynamic Fuzzy Neural Network Blind Equalization of,TN911.5
  14. Reliability Evaluation Based on Fuzzy Neural Network,S220.1
  15. Stability Based on Fuzzy Systems Design,TP11
  16. Data warehouse technology in the banking customer management systems research and implementation,TP315
  17. Association rule mining based Intrusion Detection System Research and Implementation,TP393.08
  18. Community-oriented education, personalized learning system and its implementation,TP391.6
  19. Based on data mining research tax audit case selection,F812.42
  20. Research on Clustering Algorithm Based on Genetic Algorithm and Rough Set Theory,TP18

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer software > Program design,software engineering > Programming > Database theory and systems
© 2012 www.DissertationTopic.Net  Mobile