A model reference adaptive control mechanism for controlling linear time-varying processes is introduced. In this mechanism, only one controller parameter is adjusted to compensate for variations in the plant parameters while keeping the remaining controller parameters fixed. Online estimation of the process parameters is required to find the values of the adaptive controller parameters, which can be done by allowing one controller parameter to vary while keeping the rest fixed. Then by equating the system difference equation to the model output, one can solve for the adjustable parameter. The closer the estimate of the process parameters to their true values, the closer is the convergence of the system output to that of the model. This method has a very fast convergence rate, because the controller parameter is adjusted online at every sampling instant.
展开▼
