Abstract
We present a method for self-tuning of fuzzy logic controllers based on the estimation of the optimum value of the centroids of its output fuzzy set. The method can be implemented on-line and does not require modification of membership functions and control rules. The main features of this method are: the rules are left intact to retain the operator's expertise in the FLC rule base, and the parameters that require any adjustment are identifiable in advance and their number is kept at a minimum. Therefore, the use of this method preserves the control statements in the original form. Results of simulation and actual tests show that this tuning method improves the performance of fuzzy logic controllers in following the desired reactor power level trajectories. In addition, this method demonstrates a similar improvement for power up and power down experiments, based on both simulation and actual case studies. For these experiments, the control rules for the fuzzy logic controller were derived from control statements that expressed the relationships between error, rate of error change, and duration of direction of control rod movements.
Citation Formats
Sharif Heger, A, and Alang-Rashid, N K.
Self-tuning fuzzy logic nuclear reactor controller[Proceedings of the 2nd International FLINS Workshop (Mol, Belgium, September 25-27, 1996)].
Belgium: N. p.,
1996.
Web.
Sharif Heger, A, & Alang-Rashid, N K.
Self-tuning fuzzy logic nuclear reactor controller[Proceedings of the 2nd International FLINS Workshop (Mol, Belgium, September 25-27, 1996)].
Belgium.
Sharif Heger, A, and Alang-Rashid, N K.
1996.
"Self-tuning fuzzy logic nuclear reactor controller[Proceedings of the 2nd International FLINS Workshop (Mol, Belgium, September 25-27, 1996)]."
Belgium.
@misc{etde_20030918,
title = {Self-tuning fuzzy logic nuclear reactor controller[Proceedings of the 2nd International FLINS Workshop (Mol, Belgium, September 25-27, 1996)]}
author = {Sharif Heger, A, and Alang-Rashid, N K}
abstractNote = {We present a method for self-tuning of fuzzy logic controllers based on the estimation of the optimum value of the centroids of its output fuzzy set. The method can be implemented on-line and does not require modification of membership functions and control rules. The main features of this method are: the rules are left intact to retain the operator's expertise in the FLC rule base, and the parameters that require any adjustment are identifiable in advance and their number is kept at a minimum. Therefore, the use of this method preserves the control statements in the original form. Results of simulation and actual tests show that this tuning method improves the performance of fuzzy logic controllers in following the desired reactor power level trajectories. In addition, this method demonstrates a similar improvement for power up and power down experiments, based on both simulation and actual case studies. For these experiments, the control rules for the fuzzy logic controller were derived from control statements that expressed the relationships between error, rate of error change, and duration of direction of control rod movements.}
place = {Belgium}
year = {1996}
month = {Jul}
}
title = {Self-tuning fuzzy logic nuclear reactor controller[Proceedings of the 2nd International FLINS Workshop (Mol, Belgium, September 25-27, 1996)]}
author = {Sharif Heger, A, and Alang-Rashid, N K}
abstractNote = {We present a method for self-tuning of fuzzy logic controllers based on the estimation of the optimum value of the centroids of its output fuzzy set. The method can be implemented on-line and does not require modification of membership functions and control rules. The main features of this method are: the rules are left intact to retain the operator's expertise in the FLC rule base, and the parameters that require any adjustment are identifiable in advance and their number is kept at a minimum. Therefore, the use of this method preserves the control statements in the original form. Results of simulation and actual tests show that this tuning method improves the performance of fuzzy logic controllers in following the desired reactor power level trajectories. In addition, this method demonstrates a similar improvement for power up and power down experiments, based on both simulation and actual case studies. For these experiments, the control rules for the fuzzy logic controller were derived from control statements that expressed the relationships between error, rate of error change, and duration of direction of control rod movements.}
place = {Belgium}
year = {1996}
month = {Jul}
}