In this article, the analytic optimization algorithm connected to the measurement-based predictive control framework is implemented on an exothermic tubular reactor system. The two stages of design procedure include (i) the desired input/output references, which are determined by the steady-state optimization approach, and (ii) the output regulation design of nonlinear distributed parameter systems, which is addressed using nondistributed predictive controls. Under the assumption of steady-state and plug-flow characteristics, the bang-bang type of extremal control is applied. Because of the fact that inlet perturbations could trigger the hot spots or thermal runaway, we propose two dependent manipulated variables to dominate the heat exchange function for cooling devices. With respect to specified state/input constraints, the output feedback architecture, which uses a few sensors for measurement of the reactor temperature at the prescribed axial position, is successfully demonstrated. It is a specific measurement-based feedback technique used to combat the effects of heat or unknown disturbances. All tests show that the no-offset output tracking is achieved and the undesired peak temperature is removed while physical constraints and unknown disturbances are being considered simultaneously.