MODEL-PREDICTIVE CONTROL OF A COMBINED HEAT AND POWER-PLANT USING LOCAL LINEAR-MODELS

Model predictive control has been applied to control of a combined hea t and power plant. One of the main features of this plant is that it e xhibits nonlinear process behavior due to large throughput swings. In this application, the operating window of the plant has been divided i nto a number of smaller windows in which the nonlinear process behavio r has been approximated by linear behavior. For each operating window, linear step weight models were developed from a detailed nonlinear fi rst principles model, and the model prediction is calculated based on interpolation between these linear models. The model output at each op erating point can then be calculated from four basic linear models, an d the required control action can subsequently be calculated with the standard model predictive control approach using quadratic programming .