LAUSR

Emotional-Learning based controllers are becoming increasingly popular due to their non-parametric and flexible design approach. However, most of the existing Emotional-Learning based control strategies are designed specifically for individual loops and are not suitable for superior performance in a strongly coupled MIMO system. In this technical note, a multi-variable Emotional-Learning based strategy for trajectory tracking in a strongly coupled MIMO system is proposed. This strategy incorporates an improved stimulus design that deals with tracking, regulation, disturbance-rejection and coupling effects in a systematic way. The strategy also addresses the deleterious effects of mechanical resonance associated with flexible structures from the perspective of stimulus design. The proposed strategy is validated through simulations and hardware based experiments on a 2-DOF laboratory helicopter. Effectiveness of the proposed strategy is illustrated through comparisons with an optimized multi-variable LQG controller.