Short description:
This is a project that I’ve been working with BP for more than 5 years now (and it is still on-going). My main task has been designing in-house algorithms and providing advise for modelling properties from crude and other middle distillates by using NIR spectra. The project involves lots of aspects, including multivariate exploratory analysis, signal pre-treatment, outlier detection, multivariate modelling, model optimization, non-linear regression, and many other. I’ve been also involved in developing methods to evaluate the business value of the models.
There is a non-disclosure agreement (NDA) for this project and I am not authorised to discuss any details.
Credits:
This project was developed together with BP. Several people from different teams accross the organization are involved.
Sponsors:
BP.
Presentations:
Covered by NDA agreement.
Software available:
Covered by NDA agreement.
Tags:
- Application domain: Oil & Gas
- Instrument domain: NIR, mid-IR and Raman
- Statistics domain: Continuous optimization, Multivariate calibration, Multivariate exploratory analysis, signal processing, modelling & regression