Loading...

Predictive Process Control Using a Hierachical Method Based on Regression Analysis and Artificial Neural Networks (case study: Spray Drying in Tile Industry)

Neshat, Najmeh |

482 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 39015 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Mahlooji, Hashem
  7. Abstract:
  8. This is the first attempt at process modeling in terms of predictive control using a hierachical method based on regression analysis and artificial neural networks(ANNs).This hierachical use leads to the reliability improvement of neural model of process in prediction (extrapolation and interpolation) of process output. such an outlook makes it possible to predict the proper input settings which achieved a desired process output by designing various senarios for process set up. This approach was applied in Tile industry for spray dring process and in order to indicate the achieved improvement,three models:(i) regression model of process using multiple linear regression,(ii)Neural model of process taking general approach using Multilayer Perceptron based on Back Propogation algorithm and (iii)mixed-regression neural model of process taking focus approach in architecture of neural model designed and evaluated the reliability of prediction of spray drying process output. The results indicate that the neural network models out perform the simple regression model while the mixed regression-neural model leads to the best results in prediction (extrapolation and interpolation) of spary drying process output
  9. Keywords:
  10. Predictive Control ; Artificial Neural Network ; Modeling ; Spray Dryer ; Regression Analysis

 Digital Object List

 Bookmark

No TOC