GA-BASED MODEL PREDICTIVE CONTROL OF BOILER-TURBINE SYSTEMS

This paper discusses the application of artificial intelligence based model predictive control of boiler-turbine systems. In particular, it is investigating how genetic algorithms can be used to develop an online optimal control taking the model’s nonlinear constraints such as input saturation and rate limits into consideration. It is shown that the difficulties experienced in conventional control design due to the nonlinearities and constraints can be overcome by carefully setting up the genetic algorithm and a robust control can be guaranteed over a wide range of operation.