Dissertation > Excellent graduate degree dissertation topics show

Application of Swarm Intelligence in Identification on Method of Gas-Liquid Two-Phase Flow Regime

Author: WangHong
Tutor: SunBin
School: Tohoku Electric Power University
Course: Thermal Power Engineering
Keywords: Two-phase flow Flow pattern identification Feature selection Swarm intelligence algorithm Least squares support vector machine
CLC: TK124
Type: Master's thesis
Year: 2011
Downloads: 39
Quote: 0
Read: Download Dissertation

Abstract


The gas - liquid two-phase flow is widespread in human life and industrial production processes , the phase flow pattern recognition research is of great significance . Reasonable feature selection will improve the accuracy rate of the flow pattern identification , all the characteristics of the extracted gas-liquid two-phase flow characteristics selected to study not only has a certain amount of academic significance , and modern industrial development has a very important practical value. This paper is an experimental laboratory bench to do in the air - water two-phase flow , differential pressure fluctuation signal will contain the various factors of gas-liquid two-phase flow , so the horizontal tube air - water flow differential pressure fluctuation signals collected and analyzed . First, the filtering of noise denoising collected the gas-liquid two-phase flow differential pressure signal , and then using empirical mode decomposition (EMD) and wavelet packet preprocessing methods of signal feature extraction , and thus the formation of the fusion features . Swarm intelligence algorithm for feature selection , using mainly three typical swarm intelligence algorithm , genetic algorithm ( GA ) , ant colony optimization (ACO) and discrete particle swarm optimization (BPSO) useless , redundant feature are deleted. Optimal feature subset input least squares support vector machine (LSSVM) algorithm to the variance between classes within the class and the correct classification rate as the fitness function , and then will be selected for training recognition . Four typical gas-liquid two-phase flow recognition results show that , after repeated calculations of swarm intelligence algorithm two different fitness function can eliminate redundant features reduce the computational complexity , improve the flow pattern so as to achieve the correct classification rate purposes. Comprehensive comparison of gas-liquid two-phase flow pattern identification , discrete particle swarm swarm intelligence algorithm showed overall performance than the higher of the ant colony and genetic . And experiments show that the PSO to optimize LSSVM , search speed and classification accuracy has improved further .

Related Dissertations

  1. Study of the High Temperature Gas-Liquid Two Phase Flow Rate Proportional Conrtol Valve,V233.2
  2. Study of Effect of High Temperature Seperator on Hydrodynamices in CFB Boiler Furance,TK229.66
  3. Study on Gas-Solid Two-Phase Flow in Spouted Beds,TK173
  4. Research on Text Classification Based on Biomimetic Pattern Recongnition,TP391.1
  5. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  6. Research on Feature Selection and Construction in Emotion Speech Recognition,TP18
  7. Flow Behavior Description of Two-Phase Flow Based on Image Processing Techniques,TP391.41
  8. The Design and Research of the Superheated Steam Saturator,TK17
  9. Research on Face Recognition Based on AdaBoost Algorithm,TP391.41
  10. Research on Feature Extraction, Selection and Classification Algorithms for Pulmonary CAD,TP391.41
  11. Study on Fast Detection of Gray Mold on Plants Using Spectral and Multi-Source Spectrum Imaging Technology,S432.4
  12. Coal dust movement rule and Critical Control Technology,X701.2
  13. Regeneratively cooled scramjet engine start-up phase heat transfer characteristics of,V231.1
  14. Content-based Indexing of Spam Filter Research and Implementation,TP393.098
  15. Simulation and Analysis of Diffution Rule of Somke Inside Mining Blasting Space,TD714
  16. Engine mechanical fault diagnosis system, feature extraction algorithm,TK407
  17. The Research of Moving Object Tracking Based on Swarm Intelligence Algorithm,TP391.41
  18. Research on Soft-sensor Method of Component Content in Hydrometallurgy Copper Extraction,TF811
  19. Research of Air Distribution Mode on 300MW Coal-fired Boiler,TK227.1
  20. Effects of Pipe Diameter and Dynamic Load on the Flow and Heat Transfer Characteristcs of Two-phase Boiling Flow,TK124
  21. Injection Process Simulation of Piezoelectric Device for Oil Supply Based on CFD,V233.4

CLC: > Industrial Technology > Energy and Power Engineering > Thermal engineering, heat > Thermal Engineering Theory > Heat Transfer
© 2012 www.DissertationTopic.Net  Mobile