OPTIMIZATION OF CONTROLLED VARIABLES FOR INTELLIGENT CONTROL

摘要

The paper describes an approach that uses a process model based on artificial neural networks for real-time process control and optimization. The core of the optimization algorithm is the calculation of the changes to the controlled output that is prior predicted using the artificial neural networks techniques. Detailed information concerning the structure and the parameters of the model is given in [1-2]. The target criterion for the optimization is chosen to ensure primarily the quality objectives and the stability of the controlled process. The approach is suited for processes with a large number of input and output variables overcoming the limitations of conventional process control models by allowing an adaptive control. Finally, the modeling and optimization approaches will be integrated into an existing process system for blast furnace in steel industry.