Dissertation > Excellent graduate degree dissertation topics show

Research on Identification Methods Based on T-S Model and Its Applications in Reheating Furnace

Author: SunJie
Tutor: QinShuKai
School: Northeastern University
Course: Control Theory and Control Engineering
Keywords: T-S Model Identification PSO-BP Algorithm Reheating Furnace BilletTemperature Forecast
CLC: TG307
Type: Master's thesis
Year: 2009
Downloads: 27
Quote: 0
Read: Download Dissertation

Abstract


Reheating furnace is an important thermal technology equipment in the process of billet rolling. The level of operation has great effect on the norm of quality, output, production and consumption. It provides qualified billets for hot-rolling process. However, it is difficult to measure the temperature of billet during production. So it is necessary to build a reasonable model to predict temperature.The billet’s reheating furnace is a typical complicated industry system with seriously lag-lead, multivariable, seriously coupled, time-variety, non-linear, pure time delay and a lot of disturbance factor. The billet’s temperature can not be measured online. According to the walking beam reheating furnace in this thesis, T-S model is applied to the research of the billet temperature forecasting. The model’s structure and parameters are identified apart. At first, fuzzy c-means (FCM) is used to identify antecedent parameter and structure. Then, consequent parameter is identified by least square. In order to improve the model’s precision, fuzzy-neural network based on T-S model that combines subtractive clustering and the FCM is developed in the thesis. FCM algorithm is sensitive to its initial value and liable to be trapped in a local minimum. So subtractive clustering is used to determine the initial clustering center and then FCM is used to improve the clustering’s convergence speed. Then a mixed Particle Swarm Optimizer-Back Propagation algorithm is presented to identify consequent parameter. At last, the model is simulated by MATLAB using the data gathering on-site.The simulation results have proved that the improved fuzzy-neural network based on T-S model is better in predicting the billet temperatures. If the number of rules and parameters is certain, T-S fuzzy model based on FCM and least square can identify parameter according to the only mean square error. When we want to improve the model precision, we must increase rules. However, the improved model based on PSO-BP algorithm can identify the parameters according to any mean square error in a range. This is because the mixed PSO-BP algorithm is random in essence. This model can improve the precision without increase rules, so it is more suitable for engineering. According to the temperature of billet, workers can control reheating furnace timely so that we can guarantee the quality of the billet. The research of this thesis has not only theoretical significance, but also practical application value.

Related Dissertations

  1. Research on Magnetic Field Sensor Based on GMI Effect and Geomagnetic Matching Algorithm,P318
  2. EDM Electrode Wear Mechanism and Control Research,TG661
  3. Experimental Study on Pyrolysis and Combustion Characteristics of Paint and Tar Based Single Hazardous Wastes,X705
  4. Powered Activited Carbon-UF for Treating Micro-polluted Surface Water,X703
  5. Design and Research on Stabilization Loop of Gyro Stabilized Pod Control System,V241.5
  6. Satellite Attitude Control Methods Using Only Magnetorquers,V448.222
  7. Investigation of Turbine S2 Stream Surface Direct Problem Aerodynamic Optimization Design,V235.11
  8. Temperature Drift Modeling and Compensation of Fiber-Optic Gyroscopes,V241.5
  9. The Technological Parameter Analysis and Influencing Factor Research on RUV4 Automobile Ignition Coils,U463.64
  10. The Study of Control Algorithm for Automotive Power Window Based on Lin Bus,U463.6
  11. The Research of Residential Community Design of Vanke Model,TU984.12
  12. Simulation and Analysis about Switched Reluctance Generator Power Supply System,TM31
  13. The Equivalent Circuit Model of Radial Vibration Mode Piezoelectric Transformer and Its Application,TM406
  14. Research on Compensation of Hysteresis in Piezoceramic Actuators,TM282
  15. Research on Vector Control System of Induction Motor Based on DSP,TM346
  16. The Design and Accomplishment of Simulition Platform for Voltage Clamp Experiment Based on Cell Electrophysiological Model,R318.0
  17. A Study of the Application of TBLT to English Teaching in Rural Middle School,H319
  18. Quantitative evaluation model based on information entropy of the classroom observation,G632.4
  19. High school politics,the new curriculum reform classroom teaching model to integrate exploration,G633.2
  20. Research and Implement of SDO Model in University Enroll System Based on SOA,G647
  21. Compensation Methods of Different Speech Coding for Speaker Recognition,TN912.34

CLC: > Industrial Technology > Metallurgy and Metal Craft > Metal pressure processing > General issues > Heating and heating equipment
© 2012 www.DissertationTopic.Net  Mobile