Identification of the servomechanism used for micro-displacement

Friction causes important errors in the control of small servomechanism and should be determined with precision in order to increase the system performance. This paper describes the method to identify the model parameters of a small linear drive with ball-screw. Two kinds of friction models will be applied for the servomechanism looking to rise its micro-positioning abilities. The first one includes the static, viscous and Stribeck friction with hysteresis, and the second one uses the Lugre model. The results will be compared taking into account the criterion error, the accuracy and the normalized mean-square-error of the identified mechanical parameters. The coefficients of the models are identified by a recursive identification method using data acquisition and special filtering technics. The least square identification method is used in this paper in order to establish the motor parameters used as initial condition of the recursive estimation method. Computer simulations and experimental results demonstrate the efficiency of the proposed model.