Bi-loop recursive least squares algorithm with forgetting factors

This letter develops a novel recursive least squares algorithm with forgetting factors (FRLS) named as the bi-loop recursive least squares algorithm with forgetting factors (BLFRLS). The BLFRLS is also established from the FRLS and an in-between recursive least squares algorithm with forgetting factors (IFRLS). The BLFRLS promotes the tracking capability of parameters estimation for time-varying systems without any additional data. The convergence properties in parameters estimation and predicted errors are proven. Computer simulations show that the proposed BLFRLS algorithm has better tracking capability than FRLS