Active vibration isolation control

Passive vibration isolation has a long history of development and successful implementation. The initial development was driven by the invention of the automobile and the speed increase in personal transport in general. The fact that the majority of the cars produced today still implement the technology can be seen as a testimony to its effectiveness. However, as the requirements of high-tech apparatus are ever more stringent and the accuracy levels shrink to sub-nanometer levels, the fundamental limitations of passive vibration isolation are readily insight.

This limitation is characterized by the tradeoff between the sensitivity to indirect disturbances, i.e. floor vibrations, and direct disturbances, i.e. forces that work directly on the payload. Due to this trade-off, a lot of machine time is lost by waiting on machine transients to damp out.

This tradeoff can be circumvented in two ways. The first is by adding mass to the system. This improves both the sensitivity to the direct as well as the indirect disturbances, but the effectiveness is limited by practical aspects such as floor strength and size constraints. The second method to circumvent the tradeoff is to switch to an active system. By measuring the vibrations of the floor as well as the payload, a feedforward as well as a feedback control scheme can be used.

However, the implementation of active vibration isolation schemes is not widespread and this can be attributed to a lack of understanding to the requirements for an active vibration isolation setup, as well as control schemes that are difficult to properly implement. The aim of this research is therefore to design control schemes with robust and fast implementation properties as well as optimal vibration isolation performance. The second part of the research is to develop design synthesis methods for active vibration isolation setups.

Furthermore, the full potential of active vibration isolation is the ability to compensate for active stages in the machine itself. Since non-causal filtering techniques are available for repetitive disturbances these can be canceled out up to the sensor noise levels.

project members