Observability and data validation of bilinear systems
Abstract
This paper is devoted to data reconciliation by mass balance equilibration of systems described by bilinear equations. It proposes a new approach for their resolution. We particularly insist on a decomposition algorithm which enables the problem of unmeasured variables to be treated. It allows the classification of variables into observable and unobservable, redundant and non redundant variables, and gives the relations necessary for their determination. Finally, the results of this algorithm are applied to the mass balance equilibration with the use of an iterative method based on hierarchical calculus
Origin : Files produced by the author(s)