DESIGN AND ANALYSIS OF ANN BASED CONTROLLER FOR MILK PASTEURIZATION PLANT

Abstract

The objective of this dissertation report is to study and implement the model prediction technique to identify and control linear and nonlinear process. To predict the process model an advanced technique of Artificial Intelligence i.e. Artificial Neural Network is used. Artificial Neural Network has been proved a very useful and successful tool in predicting highly nonlinear and complex processes with an impressive accuracy. The reason why Artificial Neural Network attracted researchers is due to it's capability of quickly learn the dynamics of the process, apart from that Artificial Neural Network also capable of handling the uncertain and noisy data. In this report, the plant models which are used for continually stirred tank heater and Milk Pasteurization Plant. Conventional such as PD and Artificial Neural Network controller are designed and compared with each other for continually stirred tank heater. Results show that performance of Artificial Neural Network controller is better than the conventional controller Both linear and nonlinear plant models of Milk Pasteurization Plant has been studied and represented in mathematical dynamical equation form. A supervised model prediction of Milk Pasteurization Plant has been done using Artificial Neural Network. Training has been done with three layered feed-forward network architecture with appropriate parameter adjustments and optimization technique in order to minimize Mean Square Error between reference and predicted output. Controlling of pasteurized temperature of milk plant is done with the help of generalize predictive control technique, results show that the output of the milk plant follows the reference trajectory.