Proposition of a control strategy based on the principal component analysis

The Principal Component Analysis applied to study the spectroscopic data of a reaction mixture opens a great range of possibilities to the Engineering of Chemical Reactions as reaction modelling, optimisation, monitoring and quality control of the reaction mass at real time. It is proposed a control strategy based on the similitude analysis between the abstract vectors of a reference data and the abstract vectors of the process to be controlled at real time; it is also proposed that the origin of possible deviation can be determined by means of a similitude analysis between the eigenvectors from different experiences of reference with known deviations and the eigenvectors from the deviated process.