The MULTIPLE project will strengthen EU photonics manufacturing base. Cost-efficiency, flexibility, high productivity and quality attained thanks to MULTIPLE solutions will allow laser, optical measurement and image processing equipment, companies in Europe to compete glob ally through a breakthrough technology in manufacturing, thus benefiting from business opportunities in fast growing markets.

Latest News

Share this

Laser-based IR analysers

MULTIPLE will integrate laser-based sensors for the detection of gases reaching sensitivity up to the MWIR spectral range.

Spectrometric technology for chemometric analysis has gained popularity thanks to its non-destructive nature and the capability to provide highly specific and accurate measure MWIR range are still synonyms of either sensitivity issues or expensive equipment. MULTIPLE will develop neural network models and analytical models for robust estimation of gas concentration ratios of different compounds, including CO/CO2/O2, NOx, and formaldehyde using laser-based IR spectroscopy. Different types and structures of neural networks will be studied and developed in order to reach the optimum structure for the application.

The final choice of structure will be based on number of calibrations data sets used for learning, accuracy and precision of results, immunity to cross interferences form other gas constituents in the sample, immunity to temperature and pressure variations and complexity for introduction into serial production. The neural networks will be simulated in MATLAB environment and finally deployed on the embedded spectrometer hardware platform. In parallel to development of algorithms for chemometrics, a work will be carried out towards implementation of analyzer auto calibration features which is crucial for long term autonomous operation. The auto calibration schemes will also be implemented in the embedded spectrometer platform. It is planned that a dedicated PIC microcontroller will be responsible for auto calibration and self-diagnostics functionality