Adaptive digital control of a steam turbine

The application of recent techniques of identification and adaptive control to the speed control of a steam turbine is investigated. A minicomputer was used for data acquisition, off-line and online identification and adaptive control. The turbine was modelled under different loading conditions using a recursive linear least-squares method. Although such a system is complex, second-order models were found to give reasonable results. A digital proportional-plus-integral controller with an additional feedback term was implemented to control the speed of the turbine. The controller was then made adaptive by maintaining the position of the closed-loop poles constant. This gave better performance when the turbine underwent step changes in the load. The programming required for such implementation is discussed.