Fast Adaptive Control of Time-varying Systems

Abstract For time-varying systems, the adaptive control algrithms became very complicated and the processing speed is a key problem, To speed up the computation of time-varying systems in which the parameters are usually described by time polynimials, this paper presents anew fast adaptive identification and control scheme. By means of autoregressive expansion of discrete time polynomial, the identification can be simplified to estimate coefficients of the highest power in stead of every coefficients. So that the amount of computations at each sample period is reduced greatly. Meanwhile, through the parallel processing technique, the simplified algorithm is divided into two parts which can be excuted concurrently to improve the control speed further. Finally, a simulation example which illustrates the efficiency of fast adaptive scheme is given.