On-line set-point optimization of technological processes under uncertainty is the subject of the paper for the case with important active constraints on certain process outputs. A miltilayer structure with additional upper layer follow-up constraint controller responsible for keeping the output constraints satisfied both in steady states and during transient processes is considered. The integrated system optimization and parameter estimation (ISOPE) method is investigated, i.e. iterative algorithms using an on-line steady-state feedback, designed to determine optimum operation of the plant in the presence of inaccurate models and disturbance uncertainties. A new formulation of the ISOPE method suitable for the considered structure is derived and discussed, followed by a development of an efficient dual-type algorithm. The algorithms are applied to a case study example-an industrial styrene production unit consisting of a series of distillation columns. Results of simulation runs for cases with and without errors in the feedback information from the process to the optimizer are reported.