Overview

The webinar will cover the application of two different 'machine learning' techniques. In particular, a live case study demonstration will be given on a small continuous flow reactor rig. Given very different optimisation initial conditions, the application of online 'machine learning' techniques is shown to reduce the real-life experimental effort as well as material consumption for process development. The 'Nelder Mead' optimisation technique is compared with the 'Adaptive MPC' technique, as two alternative approaches to machine learning. The initial presentation will provide a brief introduction to Nelder-Mead optimisation. Following this, the experimental rig will be introduced, with two short videos of experimental runs. Finally the results will be compared and contrasted, before an open discussion/ Q&A.

Speaker

Matthew McEwan, Perceptive Engineering Ltd

Matthew leads the formulated products Advanced Process Control team and has 15 years of experience designing and commissioning control systems with integrated analytical equipment. Matthew has a PhD in aerospace engineering from the University of Manchester.

The material presented in this webinar has not been peer-reviewed. Any opinions are the presenter’s own and do not necessarily represent those of IChemE or the Process Management and Control Special Interest Group. The information is given in good faith but without any liability on the part of IChemE.

Format

A one-hour online session: 40 minutes' presentation + 20 minutes' Q&A.
Duration (for CPD recording purposes): up to one hour.

If you are recording mandatory CPD you should refer to your own regulator's requirements as recognition of CPD hours may vary.

Webinar

This webinar is free of charge and open to all to attend, but if you wish to access the slides and a recording to replay on demand then you'll need to be a member of the Process Management & Control Special Interest Group.

If you're interested in giving an online presentation to our international community of chemical engineers, we'd like to hear from you. Please email special interest groups support team with information on yourself and your proposed talk.