Proportional-integral control optimization using imperialist competitive algorithm
PID controller is well-known for its employment in industrial automation. The applications of PID controller span from small industry to high technology industry. PID controller can be tuned using classical tuning techniques such as Iterative Methods, Direct Synthesis and Tuning Rules. However, emp...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2012
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/47510/1/FK%202012%2041R.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | PID controller is well-known for its employment in industrial automation. The applications of PID controller span from small industry to high technology industry.
PID controller can be tuned using classical tuning techniques such as Iterative Methods, Direct Synthesis and Tuning Rules. However, empirical studies have found
that these conventional tuning methods result in an unsatisfactory control performance for the nonlinear systems and most of control practitioners prefer to
tune most nonlinear systems using trial and error tuning because of this reason. A suitable tuning technique is needed for a wide range of control loops that can tune
the PID controller with minimum cost, the highest of reliability and with optimum solution.
In this dissertation, an attempt has been made to design and implement the PID Controllers by employing the Imperialist Competitive Algorithm (ICA) technique as
well as the Genetic Algorithm (GA) and the Particle Swarm Optimization (PSO) algorithm, for a selected plant. ICA is one of the newest computational algorithms that emulate the process of imperialistic competition. The system selected for modeling and simulation is a laboratory size continuous stirred tank heater (CSTH) in series with a Connecting Tank and a circulation pump.
The results from these three methods have been compared to each other based on the performance information. This comparison shows that the ICA characteristic can facilitate faster convergence to the optimal solution after 25 iterations, whereas, the GA and PSO can converge after 92 and 43 iterations respectively for the level system. Furthermore, the ICA shows the minimum cost (performance index) which is measured by the Integral of Absolute Error (IAE), in both systems compared to the GA and the PSO. The IAE from the ICA in the level system is 7.9, from the GA is
8.2 and from the PSO is 8. Moreover, the IAE values for the temperature system are 323432, 324762 and 396253 which are from ICA, PSO and GA respectively. These values show an acceptable reduction especially in the temperature system. This implies that the ICA can be used to tune the PI control loops for a continuous stirred tank heater with a minimum cost and better response in compare with the GA and
the PSO. |
|---|