Daley S., Hatonen J., Tammi K.

The “harmonic control” or “convergent control” algorithm is a well-known method to suppress harmonic disturbances in the active control of vibration. The algorithm iteratively changes the amplitude and phase of the control signal, using either the adjoint or inverse of the control path model at the disturbance frequency, so that the sum of the control path and disturbance path is driven to zero. In theory the algorithm should be implemented using a steady-state approach, where the control signal is updated only after the output of the control path has achieved steady-state in response to the previous update. In this paper an instantaneous implementation of the algorithm is proposed in which the control signal is updated at each sampling point. The benefit of this implementation is apparent: the algorithm should result in a quicker convergence rate because there is no delay associated with waiting for steady-state. To the author’s best knowledge, this paper includes for the first time convergence conditions for the instantaneous implementation in terms of constraints on the update gain and modelling uncertainty for an arbitrary square MIMO control path. These fundamental results should be of great interest to those working and researching in the area of active control of vibration. In order to validate the theoretical results, the algorithm is implemented on a “Smart Spring System” (designed and built by BAE Systems), which is a test rig that can be used to replicate typical active vibration isolation problems in Naval applications. The rig allows active control for each degree of freedom of the isolation mount, resulting in a six-input six-output system. It is demonstrated that the algorithm is capable of driving individual accelerations to zero for a harmonic disturbance. In summary, the experimental results validate the theoretical developments and illustrate the practical potential of the instantaneous implementation.