Application of optimal control system to thermal power plants is introduced. The suggested system consists of the conventional PID controllers and the control computer. It has been successfully applied to five supercritical power plants in Kyushu Electric Power Company (Total output 2,700 MW) since 1978. In the system, system identification or state space representation of the plant is performed based on the AR (Autoregressive) model describing the system dynamics. The optimal controller is designed by the orthodox Dynamic Programming procedure under a quadratic criterion function. In the paper, the procedure of the controller design and the control performance of the system are described with some results obtained both in a power plant simulation model and in the actual plants.

This content is only available via PDF.
You do not currently have access to this content.