A nonlinear model-based controller has been developed and demonstrated
for in-line neutralization of acidic mixtures. The controller consist
s of a simple, steady-state, three-parameter, nonlinear phenomenologic
al model which mimics a wide range of titration curves from strong aci
d to highly buffered systems. The model is parameterized from an-line
process data whenever nearly steady-state conditions are statistically
identified. Results show excellent control performance in the presenc
e of both ramp and step changes in acid type, acid concentration and f
low rate upsets.