Optimasi jaringan syaraf tiruan untuk kendal' daya reaktor rise'i' kartini dengan model referensi linier.
The control of the power of Kartini research reactor has been done either manually or automatically using conventional feedback controller (PID controller). In this research, an alternative control of Kartini reactor power using Artificial Neural Networks (ANN) was investigated. The research was don...
| Main Author: | |
|---|---|
| Format: | Article |
| Published: |
[Yogyakarta] : Program Magister Manajemen UGM
2001
|
| Subjects: |
| Summary: | The control of the power of Kartini research reactor has been done either manually or automatically using conventional feedback controller (PID controller). In this research, an alternative control of Kartini reactor power using Artificial Neural Networks (ANN) was investigated. The research was done using two ANNs. acting as a reactor model and a controller. The ANN-reactor model was trained using a set of input-output reactor states. The ANN-controller was trained by comparing the response of the control system and the output of a linear reference model of the reactor. The trained ANN-controller was then tested by mean of simulations. The results of the simulations showed that the ANN-controller successfully reduced the overshoot and the settling time that might occur by using a PID controller. |
|---|