Low-frequency vibration control of a pan/tilt platform with vision feedback

Abstract

This paper aims to develop active control techniques with vision feedback for suppressing low-frequency vibration of a device mounted on a pan/tilt platform due to its base disturbances. Pan/tilt platforms are commonly seen in the head of a waking robot, in an antenna, in a vision surveillance equipment, in a cannon platform, etc., for yaw and pitch direction control. However, due to its inevitable low-frequency base vibration possibly from road or sea wave disturbance in a mobile situation, orientation control of the device mounted on the top tilt platform can be seriously affected. In this paper, an adaptive sliding control (ASC) scheme is first derived and employed for vibration attenuation. Function approximation technique is used to represent the unknown disturbance in some finite linear combination of the orthogonal basis. The dynamics of pan/tilt system can thus be proved to be a stable first-order filter driven by function approximation errors. Moreover, the adaptive update law can be obtained by using the Lyapunov stability theory. Secondly, the frequently used feedback active vibration control (AVC) with filtered- x LMS algorithm is to be used and compared with the adaptive sliding control for vibration suppression performance. Experimental tests of the control algorithms show that for independent single axis excitation, about 25.14 and 23 dB attenuations in average for single-frequency disturbance have been obtained by using the ASC and feedback AVC, respectively. For dual-frequency excitation, the vibration attenuations are about 20.77 and 12.73 dB by the two methods, respectively. As for simultaneous two axes excitations, ASC and feedback AVC have respective 17.57 and 15.18 dB vibration reductions under single-frequency disturbances. Thus, validity and effectiveness of the two active control methods with vision feedback for suppressing low-frequency vibration of the device on the pan/tilt platform is verified.