Options
1990
Journal Article
Titel
Berechnung und automatische Programmierung von Zustandsschätzern für nichtlineare Prozesse mit symbolischer Datenverarbeitung
Alternative
Derivation and automatic program generation of state estimators for nonlinear processes using symbolic computation
Abstract
For the optimal control of many chemical and biochemical processes, a precise knowledge of process and product parameters which are not directly measurable on-line is required. If a sufficiently precise model of the process is available, the estimation of these unmeasurable quantities from measured data as, e.g., temperatures, pressures and concentrations, is possible based on the concepts of observers or Kalman filters for dynamical systems. As the dynamics of many chemical and biochemical processes can be described adequately only by nonlinear differential equations, the use of nonlinear observers of filters is advantageous. In this paper, a system for computer support of the calculation of extended Kalman filters using symbolic data manipulation is described. The system calculates the filter equations in symbolic form and produces optimized FORTRAN code for the simulation and the implementation of the filter. It was realized with the computer algebra system MACSYMA and fast code opt imization routines written in PASCAL.