GMC Design Based on SVM Model of Batch Reactor

Currently, SVM is an active field in artificial intelligent technology and has been applied to pattern recognition, regression and signal processing applications. It refers to kernel-based methods; follow the simple, but powerful idea of non-linearity mapping input data into high-dimensional feature space. The batch reactors are the important unit of many industries; mostly, it is used in chemical and pharmaceutical industries. In this paper, the support vector machine is employed to identify the batch reactor. The simulation results show that support vector machine provides a new and effective method for modelling a non-linear process and has promising application in industrial process. A non-linear control strategy for a batch reactor process with input-output form is proposed. This strategy is termed as internal model control (IMC). IMC is found to be well suited for a system when there is no significant rate of change of the set point.