Iterative Learning Identification for Linear Time-Varying Systems

This brief presents an approach for identifying the parameters of linear time-varying systems that repeat their trajectories. The identification is based on the concept that parameter identification results can be improved by incorporating information learned from previous executions. The learning laws for this iterative learning identification are determined through an optimization framework. The convergence analysis of the algorithm is presented along with the experimental results to demonstrate its effectiveness. The algorithm is demonstrated to be capable of simultaneously estimating rapidly varying parameters and addressing robustness to noise by adopting a time-varying design approach.