Modeling of Continually Stirred Tank Heater With ANNs Using Successive Over- Relaxation Backpropagation Algorithm  

Oleh:

Goel, A.K.

Jenis: Article from Article
Dalam koleksi: Final Program and Book of Abstracts: The 4th Asian Control Conference, September 25-27, 2002 (Sep. 2002), page 614-617.
Topik: Stirred Tank Heater; ANNs; Relaxation Backpropagation Algorithm; SOR-RPROP; CSTH
Fulltext: AC021739.PDF (82.5KB)

Isi artikel

In this paper, a quick and computationally simple gradient-descent learning algorithm, Successive Over -relaxation Resilient Backpropagation algorithm (SOR-RPROP), which is a modified version of the Resilient Backpropagation algorithm (RPROP), has been employed for training feedforward neural networks (FFNNs). It has been used to model a continually stirred tank heater (CSTH) -a MIMO process. The ANN model has been tested for large variation in the jacket flow rate (the manipulating variable), for confirming its interpolation and extrapolation capabilities. The neural net CSTH model is also tested for disturbance in the tank flow rate. It is observed that it filters out the disturbance and tracks the process nonlinear response with minimal error.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali