Model reference adaptive learning control with basis functions

Iterative learning control refers to the process of finding an appropriate control input time history to produce a desired output time history by repeated trials. We present an iterative learning control algorithm in which a reference model is used to guide the learning process. The reference model provides a means in which prior knowledge about the system (if available) is incorporated to produce a learning process with desirable transient properties. This result is the learning control counterpart of standard model reference adaptive control. In addition, basis functions are used to construct the input and output data time histories in order to reduce the dimension of the system model and to promote numerical conditioning in the learning algorithm. Experimental results on a system with highly flexible dynamics are used to illustrate the method