Control of a Benchmark Boiler Process Model with DMC and QDMC

Boiler control is an important issue for all steam generating power plants. In this paper, model predictive control was applied in a control-oriented benchmark boiler simulation model by F. Morilla (2011). In the boiler benchmark case, the advantage of model predictive approach is the ability to inherently take into account the coupling between subsystems and constraints in the system (rate constraints, variable constraints). Well-established model predictive control approaches, Dynamic Matrix Control (DMC) and the constraint-handling Quadratic DMC (QDMC) were used for SISO and MIMO boiler control cases. Controllers were built, implemented and tested with the benchmark model in Matlab and Simulink environment. The study has been used for educational purposes with M.Sc. students to give an insight of model predictive control design procedure.