Investigations into the stability of a PID-controlled micropositioning and vibration attenuation system

This paper describes a theoretical and experimental investigation into the stability and performance of an active micropositioning and vibration attenuation system, based on piezoelectric actuators incorporated into a very stiff base structure and controlled through decentralized PID feedback of the equipment position. Vibration attenuation of up to -35dB over a bandwidth of several hundred hertz was demonstrated through experiments. However, high integral feedback gain is required to achieve a sufficient vibration attenuation performance, thus inevitably creating an inherently unstable active system. Establishing guidelines concerning the design of active systems based on the described approach and identifying parameters that affect the system stability is the motivation of this research. An analytical model of a test rig was set up and validated against measurements so that the effect of structural changes on the stability of the control system could be assessed. Stiff mounts, little mass of the controlled equipment and sufficient passive damping of mounts and base were found to be essential for the stability and functionality of the active system. Based on the developed guidelines, a prototype of a fully functional micropositioning and vibration attenuation system was manufactured as part of a project aimed at building a novel magnetic resonance imaging (MRI) scanner. The maximum stable PID gains were used to assess the relative system stability since an analytical model that reliably predicted the stability margins was not available. Experiments conducted on the prototype identified and quantified the influence of various parameters on the system stability, such as the control scheme utilized, the number of controlled degrees-of-freedom, signal conditioning, the passive damping and the control loop frequency. Passive damping was considered the single most important parameter of influence.