2nd Floor, Seminar Hall, Jalvihar Guest House IIT Bombay
Introduction to the course:
The course is oriented towards Data Analytics for efficient online process operations mainly process monitoring, and quality prediction using data-driven modeling. In the drive towards good manufacturing practice in chemical plants such as oil & gas, one of the key area of focus has been leveraging data analytics based approaches for the continuous monitoring of plant performance with a view to improving plant operations. While quality can be built into the design using QbD, the need for an overall advisory system based on data analytics that continuously learns and adapts to the inherent variations during manufacturing, is being strongly felt. Such data analytics based methods are invaluable in promoting operations excellence in the batch and continuous process industry through their ability to model processes using available online data. They also play an important role in quickly detecting aberrant operation using the operating data and facilitating initiation of remedial measures so as to align the plant back along prior established superior benchmarks. Such tools are therefore proving to be very useful in the context of overall plant optimisation, advanced control, productivity enhancements & ensuring safe operations. This workshop will provide a tutorial introduction to the tools mentioned above. Representative case studies involving industrial systems taken from the batch and the continuous process industry (oil & gas, manufacturing) will be used to demonstrate the utility of the concepts presented in the course. The course will also involve hands-on sessions on representative software platforms that showcase the technology advances in information extraction and knowledge translation. Typical oil & gas related applications such as asset management, property prediction, flow assurance, and leak detection will be considered.
Pre-requisites
While the course is self-contained in terms of various statistical concepts which will be needed, familiarity with vector-matrix operations will be useful.
Learning Outcomes
1) Understanding of various data analysis based learning paradigms.
2) Use of some statistical process control techniques, and machine learning techniques for process monitoring.
3) Use of techniques such as Projection to Latent Structures, regression and Neural Networks for soft sensing.
4) Use of time series-based approaches for predictive modeling.
Click below to get the course syllabus and schedule:
(Note: All the training courses listed on this website are exclusively for nominated employees of the PSU members of CoE-OGE)
Data Analytics for Process Monitoring and Predictive Modeling