A Low-Noise Estimator of Angular Speed and Acceleration from Shaft Encoder Measurements

In the field of robotics and automation, the main objective of motion control is to guarantee accurate tracking of position or speed trajectories characterized by large harmonic content. In order to achieve this purpose, a cascade structure for the controller is usually adopted. This architecture is constituted by »nested« control loops: the outer position and speed control loops and the inner »effort« control loop [1, 2]. The outer loops mainly deal with the mechanical part of the plant, while the inner loop controls the actuator producing the effort which »moves« the system (generally an electric motor which produces torque). It is well known that the action yielded by these controllers can be split in two main parts: the feedforward one and the feedback one. The former is based on the »inversion« of the nominal dynamical model of the plant and mainly determines a fast response to reference variations. The latter is based on measurements and should make the tracking performance robust with respect to all the non-idealities, as parameter uncertainties and unexpected loads. Since the model of the controlled system and the knowledge of external loads are often quite rough, the tracking performances are essentially related to the feedback control characteristics. In order to obtain feedback controllers with wide bandwidth (i.e.