Specialized System Identification for Parafoil and Payload Systems

AIAA Aerodynamic Decelerator Systems (ADS) Conference

2013

Self Cite

The limited control authority and sensing capability of current airdrop systems make it extremely difficult for an autonomous system to recover from unanticipated changes in the wind profile near the ground. The addition of glide slope control to conventional lateral control of guided airdrop systems has shown the ability to provide dramatic improvements in landing accuracy. Additionally, upper surface aerodynamic spoilers using canopy bleed air have been shown to be an extremely effective means of both lateral and longitudinal control of parafoil and payload aircraft. The current work investigates the effect on landing accuracy of an autonomous parafoil and payload aircraft using upper surface spoilers for lateral and longitudinal control combined with a specialized control law to leverage the added authority this control mechanism provides. When used for lateral control alone, upper surface spoilers provide sufficient lateral control authority to achieve the same landing performance as typical airdrop systems utilizing the typical trailing edge brake control. Simulation results demonstrate that when using the spoilers for simultaneous lateral and longitudinal control, nearly a factor of 4 improvement is obtained in landing accuracy compared to conventional control strategies.

“…The equations of motion for this six degree of freedom parafoil and payload representation have been derived previously and validated with flight test results [14][15][16].…”

Section: Simulation Modelmentioning

confidence: 99%

Autonomous Control of Parafoils Using Upper Surface Spoilers

Ward

AIAA Aerodynamic Decelerator Systems (ADS) Conference

2013

Self Cite

“…The transformation from the body frame (frame B in Figure 7) to the Figure 7) is defined by a single axis rotation in pitch by the canopy incidence angle. The equations of motion for this six degree of freedom parafoil and payload representation have been derived previously and validated in flight testing [11][12][13].…”

Section: Figure 5: Adjusting Approach Trajectory With Lidar Wind Profilementioning

confidence: 99%

“…Gust velocities and angular rate components are computed for all three axes by driving discrete filters with unit-variance, independent white noise signals ( i  ) as shown in Eqs. (13) to (15) [14]. 1 2 …”

Section: Wind Field Modelsmentioning

confidence: 99%

Utilizing Ground-Based LIDAR for Autonomous Airdrop

Herrmann

Ward

AIAA Aerodynamic Decelerator Systems (ADS) Conference

et al. 2013

Self Cite

Uncertainty in atmospheric winds represents one of the primary sources of landing error in airdrop systems. Significant landing errors can result if the winds near the ground differ from the wind estimates used to plan the system's approach path. This work examines the use of a ground-based LIDAR system to measure the wind profile in the vicinity of the payload delivery target. The ground based LIDAR system samples the wind field at discrete points in the airspace around the target and transmits this real-time wind profile data to approaching autonomous airdrop systems. The onboard autonomous guidance algorithm is modified to incorporate the wind field information from the LIDAR to plan the approach trajectory. The performance of the autonomous airdrop system operating in conjunction with a ground-based LIDAR is evaluated in simulation. A 3-dimensional, time varying wind model is used to generate realistic wind conditions that constantly vary over the course of the vehicle's trajectory and the parafoil and payload is simulated using a 6 degree of freedom non-linear model. Monte Carlo simulations are used to generate landing accuracy statistics to evaluate the improvement in landing accuracy when the information from the LIDAR unit is incorporated into an otherwise typical guided airdrop system. Simulation results demonstrate a factor of three improvement in landing accuracy in the presence of wind shear when incorporating the wind information from the LIDAR. Results also demonstrate that wind measurements covering the final approach phase of the flight are most critical and that it is possible to obtain large improvements in landing accuracy with even a small number of discrete measurements of the wind profile near the ground.

“…In the numerical simulation, the package is modeled as a rigid body with 6 degrees of freedom, which is typical in air vehicle dynamic modeling [6]. The degrees of freedom include the three components of the body's mass center position with respect to an inertial reference frame and the four quaternion parameters representing the body's orientation.…”

Section: Iiia Flight Dynamic Modelmentioning

confidence: 99%

Design, Simulation, and Experimental Testing of Humanitarian Aid Airdrop Micro Packages

Herrmann

AIAA Atmospheric Flight Mechanics Conference

Montalvo

et al. 2012

Self Cite

Conventional airdrop methods for humanitarian aid and emergency relief require dropping heavy payloads far away from the intended recipients. Currently, there are a number of issues with this method of delivery. Because supplies are distributed by just a few large crates dropped on or near the location of those requiring relief, there is not only risk of injury upon being struck by one of these large crates, but it is also common for the distribution of supplies to be inhibited by adversaries. Thus, in an effort to reduce or eliminate the occurrence of such difficulties, a safer, more effective means of dropping and distributing humanitarian aid is necessary. The development of small sized packages of supplies that can be dispersed directly above a populated area with minimal risk of human injury is essential to ensure safe, timely, and effective distribution of aid. The research reported here explores several proposed packaging designs of individual food and water rations to identify possible methods to deploy emergency relief in a manner consistent with the previously stated requirements. With a combination of flight dynamic simulation, wind tunnel testing, and flight testing, promising package designs that meet impact requirements are identified. NomenclatureC = aerodynamic force or moment coefficient, denoted by subscript , , = unit vectors denoting the axes of a reference frame [I] = total body mass moment of inertia matrix (kg-m 2 ) A = reference area (m 2 ) F s = cushion spring force (N) g = gravitational acceleration (m/s 2 ) k s = cushion spring stiffness (Pa/m) L, M, N = moment components in the body reference frame (N-m) m= total body mass (kg) p, q, r = angular velocity components of the body frame with respect to the body frame (rsd/s) q 0 , q 1 , q 2 , q 3 = quaternion parameters denoting the orientation of the body frame relative to the inertial frame R = sphere radius (m) r x , r y , r z = body frame components of a position vector from center of gravity to a point of interest (m) s = cushion thickness (m) SL, BL, WL = distances along the station, butt, and water lines from the body frame origin (m) [T BW ] = rotation matrix from the body to the wind reference frame [T IB ] = rotation matrix from the inertial to the body reference frame 2 u, v, w = velocity components of the mass center in a body reference frame (m/s) V air = air velocity magnitude relative to the body center of gravity (m/s) V imp = cushion impact velocity magnitude (m/s) V w = horizontal wind velocity magnitude with respect to the inertial frame (m/s) x, y, z = inertial positions of the body mass center (m) X, Y, Z = force components in the body reference frame (N) x imp = cushion impact deflection (m) φ, θ, ψ = Euler roll, pitch, and yaw angles of the body frame relative to the inertial frame (rad) ψ w = wind heading from North (rad) Subscripts * 0 = pertaining to the initial state * A = aerodynamic forces and moments acting on the attachments * air = air velocities relative to the center of gravity in the body frame * B