Loading...

Multiple-Model Predictive Control for a Fed-Batch Bioreactor of Ethanol Production

Rabbani Mehr, Alireza | 2024

4 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56833 (06)
  4. University: Sharif University of Technology
  5. Department: Chemical and Petroleum Engineering
  6. Advisor(s): Pishvaie, Mahmoud Reza
  7. Abstract:
  8. Bioethanol is an alcohol that is produced by microbial fermentation from carbohydrates produced in sugar-containing plants such as corn, sugarcane, or lignocellulosic biomass. Bioethanol is mostly used as an additive to gasoline. Cars can run on a 15% blend of bioethanol and gasoline and need to be modified to use pure bioethanol. Ethanol in vehicles is beneficial for the environment Because ethanol pollutants are cleaner than greenhouse gases. Ethanol is also used to produce ethyl acetate, which is a widely used solvent in the printing and packaging industries. In this research, fed-batch bioethanol reactor was investigated. First, the modeling of the bioreactor was done and the behavior of the system was analyzed in the open loop mode with changes in the system. These results show the nonlinearity of the system and the change in gain with different operating points. Then, model predictive control algorithms were designed without offset to remove unwanted fluctuations. This controller is based on multiple models, which can switch between offline models with the range of the system's operating point. The switch system is from one model to another using offline categories. The center of each cluster is representative of the operating points of that operational range. This controller operates on multiple inputs and multiple outputs and the manipulated variables are the substrate and reactor temperature. To control the system, the flow rate of the input to the reactor and the flow rate of the jacket fluid are controllers. The system was investigated as a single-model and multi-model in the presence of uncertainty and disturbance by applying constraints on the controller. The results show that the multi-model controller system has a much better performance in terms of speed and accuracy and is more resilient against entering the system. MATLAB software was used for data classification and reactor simulation and control
  9. Keywords:
  10. Multimodel Predictive Control (MMPC) ; Bioethanol ; Semi Continuous Method ; Multiobjective Optimization ; Observer Design ; Semi-Continuous Reactor Control ; Bioethanol Reactor Modeling

 Digital Object List

 Bookmark

No TOC