Title :
On-line control model validation through simulation
Author :
Koskinen, J. ; Ritala, R.
Author_Institution :
Finnish Pulp and Paper Res. Inst., Espoo, Finland
Abstract :
The use of digital signal processors (DSP) in the evaluation of control models and in the detection of process and sensor faults is discussed. Based on measured process inputs a DSP can be used to simulate the process. A multivariate statistical analysis is applied on the difference between the true outputs and the simulated outputs. The detection and identification methods for model invalidity are discussed. Some important pulp and paper applications are outlined
Keywords :
control system analysis computing; digital signal processing chips; digital simulation; fault location; identification; real-time systems; statistical analysis; digital signal processors; digital simulation; identification; model invalidity; multichannel regressive; multivariate statistical analysis; online control model validation; paper industry; pulp industry; sensor fault detection; Adaptation model; Digital signal processing; Failure analysis; Fault detection; Fault diagnosis; Job shop scheduling; Predictive control; Predictive models; Signal processing; Workstations;
Conference_Titel :
Control Applications, 1993., Second IEEE Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-1872-2
DOI :
10.1109/CCA.1993.348265
