Optimized neurocontroller for vibration reduction in helicopter blades

High levels of vibrations in helicopters contribute to fatigue damage in structural members, degraded ride-quality, and performance problems in on-board systems. Both passive and active means of vibration control have been studied in this context. The former approach involves either the use of passive devices or an optimal tailoring of the mass and stiffness properties of the structure to minimize vibrations under anticipated loading conditions. The active approach involves the use of hub-mounted control systems or actively actuated control surfaces on the blade to vary the airflow over the blade, thereby limiting the vibration levels. Each of these approaches, in particular the latter, requires that an accurate model for predicting the dynamic response of the blade be available. Analytical models available for predicting dynamic response of helicopters are somewhat inadequate, and give rise to uncertainties that must be taken into account when selecting control strategies. Linear models based on system identification are ineffective due to the inherently nonlinear nature of the system and to the presence of large uncertainties. The present paper proposes the use of artificial neural networks, with on-line learning capabilities, to develop robust controllers for this class of dynamic systems. The paper focuses on the design of such an adaptive neurocontroller, performed in conjunction with a multidisciplinary design of the rotor blade for optimizing system response.