Fw: Workshop on Measurement-based Optimization in the Presence of Uncertainty

[Raghunathan Rengasamy]

 

---

From: Sridharakumar Narasimhan
Sent: 09 March 2019 09:52
To: Raghunathan Rengasamy; Shankar Narasimhan S; Arun K Tangirala; Niket Kaisare; Rajagopalan Srinivasan; Ramkrishna Pasumarthy; Arunkumar D Mahindrakar; Rachel; Bharath Bhikkaji; Dr. Prakash Jagadeesan
Subject: Fw: Workshop on Measurement-based Optimization in the Presence of Uncertainty

 

Dear all,

Apologies for cross posting. Please pass on this announcement to those interested. 

 

Sridharakumar Narasimhan,

Dept. of Chemical Engineering,

IIT Madras,

Chennai, India, 600036

Web: 

www.che.iitm.ac.in/~sridharkrn

---

From: Sridharakumar Narasimhan
Sent: Thursday, March 7, 2019 5:23 PM
To: fac...@rbcdsai.org; gr...@rbcdsai.org
Subject: Workshop on Measurement-based Optimization in the Presence of Uncertainty

 

Dear all,

Workshop on Measurement-based Optimization in the Presence of Uncertainty

• Venue : RBCDSAI, 5th floor, Biotechnology Building 2.
• Day-1 : 02:00 PM to 05:30 PM on 15th March 2019
• Day-2 : 09:00 AM to 05:30 PM on 16th March 2019

The Robert Bosch Center for Data Sciences and Artificial Intelligence will be hosting a 2-day workshop on "Measurement-based Optimization in the Presence of Uncertainty".

The speakers are

• Prof. Dominique Bonvin, EPFL, Switzerland.
  Currently Chevron visiting professor in Dept. of Chemical Engineering, IIT Madras.
• Prof. Shankar Narasimhan and Dr.Sridharakumar Narasimhan,
  Dept. of Chemical Engineering, IIT Madras.

  Brief Abstract: The workshop will deal with measurement-based real-time optimization (RTO) strategies for improving process performance in the presence of uncertainty in the form of plant-model mismatch, drifts and disturbances. RTO typically uses a plant model to compute optimal inputs.

  In the presence of uncertainty, selected model parameters can be estimated and the updated model then used for optimization. Although very intuitive, this two-step approach suffers from the fact that the model is almost invariably “inadequate”, which prevents from reaching the plant optimum. Other approaches have been developed in the last two decades to overcome this difficulty. For example, in modifier adaptation, the basic idea is to leave the model parameters unchanged but to use plant measurements to “appropriately” modify the cost and constraint functions. Also, researchers have tried to turn the optimization problem into a control problem, as this is done in self-optimizing control.

  We will address several important issues such as (i) what can be done off-line prior to process operation, and what should be performed in real time, (ii) how much of the optimization effort is model-based and how much is data-driven, (iii) what to measure, what to adapt, how to adapt?. 
  
  

  Who should attend: Graduate students, senior undergraduate students, researchers interested in optimal control, dynamic optimization, real-time optimization. Registration is mandatory. Please registe​r at 
  

  
  

https://docs.google.com/forms/d/e/1FAIpQLSfFeG8amTBNZhDM7KfNBp0JIqJfdHKmU8q9cCd5aSZBWx4guQ/viewform​

Sridharakumar Narasimhan,

Dept. of Chemical Engineering,

IIT Madras,

Chennai, India, 600036

Web: 

www.che.iitm.ac.in/~sridharkrn

best student paper nominations at AAMAS.