Experimental Investigation of Real-Time Helicopter Weight Estimation

M ASS properties are critical parameters that influence stability and handling qualities of any flight vehicle. Real-time knowledge of mass properties is especially useful for helicopter flight, where payload release or acquisition may routinely alter the weight or mass center location in an unknown manner. Examples include a helicopter drawing up water to fight forest fires, a medivac helicopter picking up injured individuals, or a combat helicopter dropping off or picking up supplies or troops.Advanced flight control systems can use such information to ensure safe operation of the helicopter under various flight regimes and to provide feedback for gain scheduling. Additionally, knowledge of the mass properties during flight can reduce maintenance costs by helping to ascertain when life-limited parts need to be inspected or replaced in a precise way. Previous work has been conducted to estimate gross weight of helicopters during flight. Methods include hover performance charts [1], neural networks [2–4], and corrected moment theory [5]. Most recently, Abraham and Costello [6] successfully estimated weight and mass center of a helicopter in a simulation environment by making use of an extended Kalman filter. The purpose of this note is to apply the methods of Abraham and Costello to experimentally estimate the gross weight of a small radio-controlled (R/C) helicopter, the Align TREX600e. A nonlinear dynamic helicopter model is developed, and main rotor lift curve slope is estimated using maximum likelihood estimation. Flight tests for various helicopter weights are conducted, and it is shown that the gross weight can be reliably estimated in real time during axial maneuvers. The estimator is also shown to be responsive to instantaneous changes to the helicopter weight during flight.