Iterative learning identification modeling of second-order linear time-variant models for ultrasonic motors
To improve accuracy of the speed control model for ultrasonic motors ( USM),a second-order linear time-varying model based on iterative learning identification ( ILI) was proposed.A linear time-va-rying model was established with the driving voltage frequency as the input variable and the rotational speed as the output variable to capture the dynamic behavior of the USM,addressing its nonlinear and time-varying characteristics.An enhanced ILI algorithm was developed by optimizing an objective func-tion that combines error-weighted norms and parameter variation rate norms,and a new learning law was designed with a weight matrix constructed using a normalized Cauchy filtering window.Simulation and ex-perimental results demonstrate that the proposed second-order linear time-varying model accurately de-scribes the speed dynamics of the USM,while the ILI algorithm exhibits excellent convergence and ro-bustness,effectively addressing time-varying disturbances and improving the motion control performance of the USM.
ultrasonic motorlinear time-variantrite systemiterative learning identificationlearning lawtwo-norm minimizationspeed control model
NETL
NSTL
万方数据
