Related Experiment Videos
Output feedback neurolinearization
R B Boozarjomehry1, W Y Svrcek
1Chemical and Petroleum Engineering Department, The University of Calgary, Canada.
ISA Transactions
|May 23, 2001
Summary
A novel output-feedback neurolinearization method achieves model-independent input-output linearization. This advanced control strategy outperforms traditional methods in setpoint tracking and disturbance rejection for chemical processes.
Area of Science:
- Chemical Engineering
- Control Systems
- Artificial Intelligence
Background:
- Model-independent control is crucial for complex industrial processes.
- Existing linearization techniques often require accurate process models.
Purpose of the Study:
- To introduce and evaluate a new model-independent control method: output-feedback neurolinearization.
- To compare its performance against Global Linearizing Control (GLC) and PI controllers.
Main Methods:
- Developed an output-feedback neurolinearization technique utilizing only system input-output data.
- Applied and compared the method to temperature control in a CSTR reactor and pH control in a neutralization process.
Main Results:
- Output-feedback neurolinearization demonstrated superior performance in setpoint tracking.
- The method showed enhanced disturbance rejection capabilities compared to GLC and PI controllers.
- The technique's model-independent nature was confirmed as a significant advantage.
Conclusions:
- Output-feedback neurolinearization offers a robust and effective alternative for process control.
- Its model-independent nature simplifies implementation and broadens applicability.
- This method holds significant potential for improving control performance in various chemical engineering applications.