System Identification of Batch Reactor Using Machine Learning Techniques

This aim to develop the Machine Learning models to predict the temperature of a batch reactor using machine learning techniques. Various machine learning techniques like Linear Regression, Decision Tree model, Random Forest, and Support Vector Machine are used to model the system. The open loop data of reactor is used to predict the best model and evaluate the performance of these models in predicting the temperature of the reactor. The comparison of each model and select the most precise model for predicting the temperature of the batch reactor. The proposed approach has the ability to significantly improve the model accuracy and accurate prediction of the reactor temperature, which can lead to more effective process control and optimization. As a case study, the input–output data of the highly nonlinear batch reactor is considered for the model fit. The machine learning models can be further used for the predictive controller design for validating on an experimental setup. Further, these models will be used for the Nonlinear Model Predictive Controller (NMPC) design via Python and validation using Jetson Orin Nano board.