Experimental Evaluation of Adaptive Control Methods on a Pilot Plant Evaporator

Three adaptive control methods (STC, APCS and SFC) based on different combinations of estimation laws and controller design methods are experimentally evaluated by application to the computer-controlled pilot-plant evaporator at the University of Alberta. The three methods are also analyzed with respect to the performance of the parameter estimation law (recursive least squares versus the projection algorithm) and controller design objectives. The overall conclusions observed here cannot be generalized to the point where one can say that one controller is always better than the other. However, it is demonstrated that adaptive control can produce improved performance in practical applications and that each application is unique and requires careful analysis and planning by the control engineer during the implementation and operation stage.