Application of exploratory modelling to active headrest control

A new technique of iterative modelling and controller tuning is presented and its applicability demonstrated on a model of an active headrest which attenuates the ambient noise for the comfort of a passenger. The underlying scheme relies on internal model control (IMC) which is shown to have self-excitation for identification for control. Application of IMC creates synergy of identification and control and models tuned for control can bring enhanced and more reliable performance. The allowed gain error (AGE) is used which is compared with model uncertainty to form a criterion of modelling for control. This modelling criterion quantifies the maximum violation of robust stability and performance robustness along the frequency axes, as such it is a criterion of modelling for control. The new scheme is called exploratory modelling for controller optimization (EMCO) and its objective is to obtain simple robust controllers. The EMCO scheme can be used both on-and off-line.