Modeling and controller design for a Conical tank process using Radial Basis Function neural network

EasyChair Preprint no. 402

Abstract

The major issue in the process industry is to control the level of the liquid in the nonlinear processes especially in a conical tank. This study addresses the efficient controller design for the conical tank process. The conical tank is divided into different operating zones and the approximated first order process model (FOPDT) was identified for each region using simple black box technique. A neural network model based on radial basis fucntion (NNRBF) was found from the FOPDT model. For the identified network, predictive controller (NMPC) was proposed using NNRBF. The performance indices and time domain specifications of the proposed controller are compared with the conventional direct synthesis PI controller (DSPI) and internal model controller (SIMC). The result showed that proposed predictive controller is more effective and robust compared with the other controllers.

Keyphrases: Conical tank, Neural Model Predictive Control, PRBS, Radial Basis Function

@Booklet{EasyChair:402,
  author = {Venkatesh Sivanandam and Ramkumar Kannan},
  title = {Modeling and controller design for a Conical tank process using Radial Basis Function neural network},
  howpublished = {EasyChair Preprint no. 402},

  year = {EasyChair, 2018}}