This will be a 1 hour Zoom webinar as part of the RSC Process 91AV and Technology group's monthly webinar programme.
Spectroscopic measurements used for process monitoring and control often rely on multivariate mathematical methods to enable accurate estimation of the concentrations of specific chemicals from complex mixtures. These quantitative models are time and resource expensive to build and maintain. Furthermore, mathematical models may be affected by changes in physical parameters, such as changes in temperature or the number and size of particles present (e.g., crystallisation). These variations introduce errors in the estimation of chemical concentrations because of their effects on the measured spectra. Another issue that may occur is the need to transfer or update a calibration model because of changes in the process or changes in the spectrometer instrumentation used. For example, when deploying a model developed in the laboratory into a process environment. This webinar will illustrate advanced algorithms devised to alleviate these issues using several process examples.
Spectroscopic measurements used for process monitoring and control often rely on multivariate mathematical methods to enable accurate estimation of the concentrations of specific chemicals from complex mixtures. These quantitative models are time and resource expensive to build and maintain. Furthermore, mathematical models may be affected by changes in physical parameters, such as changes in temperature or the number and size of particles present (e.g., crystallisation). These variations introduce errors in the estimation of chemical concentrations because of their effects on the measured spectra. Another issue that may occur is the need to transfer or update a calibration model because of changes in the process or changes in the spectrometer instrumentation used. For example, when deploying a model developed in the laboratory into a process environment. This webinar will illustrate advanced algorithms devised to alleviate these issues using several process examples.
