Aerial cargo delivery, also known as airdrop, systems are heavily affected by atmospheric wind conditions. Guided airdrop systems typically employ onboard wind velocity estimation methods to predict the wind in real time as the systems descend, but these methods provide no foresight of the winds near the ground. Unexpected ground winds can result in large errors in landing location, and they can even lead to damage or complete loss of the cargo if the system impacts the ground while traveling downwind. This paper reports on a ground-based mechatronic system consisting of a cup and vane anemometer coupled to a guided airdrop system through a wireless transceiver. The guidance logic running on the airdrop system's onboard autopilot is modified to integrate the anemometer measurements at ground level near the intended landing zone with onboard wind estimates to provide an improved, real-time estimate of the wind profile. The concept was first developed in the framework of a rigorous simulation model and then validated in the flight test. Both simulation and subsequent flight tests with the prototype system demonstrate reductions in the landing position error by more than 30% as well as a complete elimination of potentially dangerous downwind landings.

No abstract

No abstract

Precision placement of guided airdrop systems necessarily requires some mechanism enabling effective directional control of the vehicle. Often this mechanism is realized through asymmetric deflection of the parafoil canopy trailingedge brakes. In contrast to conventional trailing-edge deflection used primarily for lateral steering, upper-surface bleed air spoilers have been shown to be extremely effective for both lateral and longitudinal (i.e., glide slope) control of parafoil and payload systems. Bleed air spoilers operate by opening and closing several spanwise slits in the upper surface of the parafoil canopy, thus creating a virtual spoiler from the stream of expelled ram air. The work reported here considers the autonomous landing performance of a small-scale parafoil and payload system using upper-surface bleed air spoilers exclusively for both lateral steering and glide slope control. Landing accuracy statistics computed from a series of Monte Carlo simulations in a variety of atmospheric conditions and experimental flight tests were found to be in good agreement. Median miss distances for the combined lateral and longitudinal control logic are on the order of 13 m, indicating an improvement in landing accuracy of nearly 50% over similar systems employing only lateral steering control. Nomenclature F = vehicle turn rate mapping GS min , GS max = minimum and maximum system glide slope GS c = commanded system glide slope H = system glide slope mapping h = current altitude, m I I , J I = inertial reference frame axes along the north and east directions I wf , J wf = wind-fixed reference frame axes along the downwind and crosswind directions L = instantaneous distance from target, m R = turn radius, m V 0 = estimated horizontal projection of vehicle airspeed, m∕ŝ V W;x ,V W;y = estimated wind velocity components along inertial north and east directions, m∕s x, y = inertial position components of system mass center, m x T , y T = target coordinates, m x wf , y wf = wind-fixed components of system mass center, m _ z = system sink rate, m∕s α = aerodynamic angle of attack, rad β = sideslip angle, rad δa, δs = asymmetric, symmetric spoiler deflection δl, δr = left, right spoiler deflection χ 0 = system total velocity azimuthal angle, rad ψ = system heading angle, rad ψ c = commanded system heading angle, rad ψ W = wind direction, rad

Advances in guided airdrop technology including guidance, navigation, and control algorithms, novel control mechanisms and wind sensing algorithms have led to significant improvements over unguided airdrop systems. Guided systems are autonomously controlled with an embedded microprocessor using position and velocity feedback. While capable of highly accurate landing, these systems struggle to overcome deviations from expected flight dynamics due to canopy damage or cargo imbalance, complex terrain at the drop zone, and loss of sensor feedback. Human operators are intelligent, highly adaptive, and can innately judge the flight vehicle and environment to steer the vehicle to the desired impact point provided sufficient information. This work experimentally explores operators' abilities to accurately land an airdrop system using different sensing modalities. Human operator landing results are compared with a state of the art fully autonomous airdrop system. Across the methods analyzed, human operators attained up to a 40% increase in landing accuracy over the fully autonomous control algorithm.