To cope with the problem of nonlinearity and/or time-varying process dynamics in the presence of unmeasurable disturbances, a single input-single output (SISO) adaptive inferential control (AIC) system based on an adaptive deconvolution model is proposed. Furthermore, a cautious on-line parameter estimation method and a predictive model for disturbances are included in the AIC to improve its ability of rejecting disturbances. The design of the SISO AIC is also extended to a multivariable control system for control of a multiple input-multiple output (MIMO) process. Finally, the AIC is applied to a process with limited measurement problems. Computer simulations are used to demonstrate the utility of the AIC. |