Investigating the implementation of Recurrent Neural Network using the Lyapunov stability learning algorithm
Date:
The aim of this research is to empirically realize some chemical engineering systems using an artificial network trained with a view to using it in an adaptive model-based predictive control scheme. Among other things,the research intends to:
i. Implement a Gated Recurrent Unit Neural Network using the Lyapunov stability learning algorithm.
ii. Model and simulate some chemical engineering processes using mathematical tools.
iii. Identification of the processes in (ii) above using the neural network model developed in (i) above.
iv. Implemeting a Neural Network MPC scheme on the processes in ii) above using the model realized in iii) above.
The proposed research will provide insight into how chemical engineering processes can benefit from emerging trends in the field of Machine learning (ML). It will ultimately provide a better understanding on how state-of-the-art neural network architecture can be integrated with more efficient mathematical technique and be employed in chemical process industry to fine tune their control scheme and thus realized a set goal amist ever changing constraints.