Title :
Performance assessment of DMC controllers
Author_Institution :
Control Arts, Danville, CA, USA
Abstract :
The theory and technology of dynamic matrix control (DMC) or multivariable controllers is well developed, and these controllers have been running in industrial applications for several years. There is heightened interest in assessing the performance of these controllers. The multivariable performance measurements developed so far are computationally intensive, often require complete transfer functions, and are strictly speaking for the unconstrained case. To a certain extent however, the single loop metrics are applicable for the multivariable case, and are at any rate an improvement over using set-point error variance alone. One of the most common metrics is the Harris performance index. The paper discusses if performance indices are suitable for the evaluation of DMC controllers. As demonstrated, one of the key assumptions of many performance indices is not realized for DMC control, and modifications must be made if these indices are to be applicable
Keywords :
autoregressive moving average processes; controllers; industrial control; multivariable control systems; performance index; ARIMA model; Harris performance index; dynamic matrix control; industrial control; multivariable controllers; single loop metrics; Degradation; Error correction; Feedback control; Industrial control; Monte Carlo methods; Performance analysis; Predictive models; Servomechanisms; Three-term control; Transfer functions;
Conference_Titel :
American Control Conference, 1999. Proceedings of the 1999
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-4990-3
DOI :
10.1109/ACC.1999.786548