Robot Experiment Analysis of Airport Ramp Area Time Constraints

Abstract

Improvement in airport operations using optimization schemes has been an active research area in the recent years. Particular attention has been given to improve taxiway and runway queue operations. However, once these operations are improved by an efficient taxiway schedule, its execution relies on the planning of ramp-area aircraft movements. An important step in the integration of the taxiway schedule with the planning of ramp-area aircraft maneuvers is to understand the constraints imposed on the aircraft trajectories due to the geometry of ramp-area and aircraft kinematics. Data for ramp trajectories are usually unavailable. To address this, we use an inexpensive scaled-down robot experiment to collect some critical data about aircraft trajectories. Ramp movement trajectories are then modeled by stochastic processes since they are heavily dependent on the human operator. We use the stochastic model to analyze the relationship between aircraft pushback time intervals and ramp-area conflicts. We then discuss constraints that can be imposed on aircraft pushback intervals to avoid any conflicts among trajectories.

A Mixed Integer Linear Program for Real-Time Computing the Optimal Push Back Time Windows

“…The program defined by objective (16) and constraints (17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33) solves for the optimal time window that is constrained to not intersect the conflict quadrilateral Q. For every additional conflict quadrilateral, the nine constraints (25)(26)(27)(28)(29)(30)(31)(32)(33) are required to separate the time window.…”

“…Modeling ramp area aircraft departure maneuvers as stochastic processes, we sample a large number of departing trajectories from the stochastic model [16,17]. The sampled trajectories define a family i of feasible trajectories that originate from a single gate and arrive at the target movement area at the scheduled Figure 1.…”

“…1c. The distribution in trajectory duration is estimated from the robot experiment data which is directly influenced by the human operator [16,17]. We use data from a scaled down robot experiment of the ramp area because trajectory data are not readily available mainly due to the lack of surveillance equipment in the ramp area.…”

“…To address the uncertainty, a stochastic model of aircraft trajectories [16,17] was proposed. The model was used to sample a large number of feasible ramp area aircraft trajectories within the ramp area.…”

“…Using the sampled trajectories, we computed conflict distributions among any two aircraft defined by their relative target movement area schedule [16,17]. The conflict distributions are defined by a measure contain zero conflict points.…”