For realizing the yaw-heave coupling control of unmanned helicopters in full flight envelope, an active modeling based controller is developed based on robust model predictive control and adaptive set-membership filter estimation. A steady increment based model predictive control is proposed, which makes use of the estimated values and boundaries of full states and model errors to deal with the varying yaw-heave dynamics and actuator delay in full flight envelope. This scheme also results in lower computation times and the ability to satisfy the input constraints with partial state information for effective helicopter control. Using the identified hovering and cruising dynamics model, real flight experiment is done to test the performance of the proposed controller in full flight envelope on our UAV platform. The results demonstrate that the active modeling based controller can solve the actuator delay and varying yaw-heave dynamics, and is feasible for the yaw-heave control of unmanned helicopters in full flight envelope.