Distributed control

In the context of flexible operation of integrated processes, we are working on distributed control architecture due to its ability to capture key system interactions and practical implementation capability. Decomposition of a large integrated network into distributed subsystems is an important and open issue in the design and implementation of distributed controllers. The key challenge in this decomposition problem is the multiplicity of performance objectives. We are working on this problem using a novel approach of community detection. Specifically, we have been able to solve the decomposition problem for two performance objectives; minimum inter-controller communication and maximum input-output impact. Our collaborators have shown the effectiveness of our solutions through rigorous simulations. We are working on extending this work to simultaneous estimation and control, big-data systems and other performance objectives.

For more informations, you can refer to our published work.

This work received Prof Krithi Ramamritham award for creative research.