Parallel Kalman filter networks for kinetic methods of analysis

✍ Scribed by Peter D. Wentzell; Stephen J. Vanslyke

• Publisher: Elsevier Science
• Year: 1992
• Tongue: English
• Weight: 780 KB
• Volume: 257
• Category: Article
• ISSN: 0003-2670
• DOI: 10.1016/0003-2670(92)80167-6

✦ Synopsis

AbStiCt An algonthm for the correctIon of errors ansmg from between-sample vanatlons m the pseudo-first-order rate constant for kmetlc methods IS dermbed The algonthm 1s based on the hnear Kalman filter and uses a set of models (typically 40 or more) m parallel The best model for a given set of data IS selected by exammatlon of the mnovation sequences Lmutattons of this approach with respect to data range, number of data points and measurement noise were mvestlgated with simulated data ExperImental data from the molybdenum blue method for the determmatlon of phosphate were used to demonstrate the msensltlvlty of the algonthm to vanatlons m the rate constant Prmc~pal advantages of the new method over other error-compensatron methods are Its srmphaty, adaptabdlty, stabdlty, fmed cycle time and memory reqmrements, and Its ablhty to provide results m only one pass through the data The method can be used m real time and IS well suited for dlgltal slgnal processors and parallel computmg architectures Keywords Kmetrc analysts, Error correction, Kalman filter