A route selection problem applied to auto piloted aircraft pushback tractor M. Cassaro and G. Sirigu Politecnico di Torino, Turin, Italy Abstract. Airplanes taxiing on taxiways in airports burn a large amount of fuel, emit tons of CO2, and are very noisy. The aviation industry demands alternative means to tow airplanes from gate to take-off with engines stopped (dispatch towing). Pushback tractors are the main candidate to accomplish this mission. The major issues that must be solved to make this solution applicable are mainly two: structural and regulatory limitations. From a structural point of view, new technologies are already under investigation to optimize the towbarless tractors joint and drastically reduce the fatigue loads on the nose landing gear (NLG). From a regulation point of view, to guarantee safety during towing an automatic control system for the tractors, capable to control taxiing speed and route inside the airport area, will be the solution. In this paper, we present a software solution for a route selection problem in a discretized airport environment. The algorithm, implemented using Hopfield neural network, is able to compute the shortest allowed path in terms of checkpoints. These, once passed to the tractor autopilot, should be able to guide it from the airport tractor parking zone to the selected parked aircraft (phase 1), perform taxiing (phase 2) and going back from the runway threshold to the parking area (phase 3). The phases are in reverse order when landing occurs. 1 Introduction A quite significant revolution is currently undergoing in the aircraft ground operations system. The taxi phase has always been critical for several reasons such as fuel consumption, noise, pollution and foreign object damage (FOD). All these issues could be resolved all in once by having the possibility of keeping the engines off during ground maneuvers. Therefore, a new way of thinking at the pushback tractor is raising in the aeronautical industries. Instead of using them only to exit the parking lot, allow the aircraft towing from gate to take-off position and from landing arrival point to the gate. This new usage of the pushback tractor introduces some problematic from structural and regulatory point of view. Structurally NLG are not designed to be towed for long distances, which fact can lead to fatigue failure due to transversal loads. To this purpose, recently patented towbarless system has been designed, built and tested by an industry consortium to avoid solid link between the tractor and the landing gear structure. The other critical issue is about the aviation regulation, since aircraft not having a Pilot in Control (PIC), when towed by the tractor driver, face safety, responsibility and regulatory limitations. Our research team is working on this last aspect by designing a novel concept of airport ground operation system [1]. It consists in a semi-autonomous system, activated by the control tower, in which autonomous auto piloted tractors are capable of accomplish towing missions between points selected by the tower operators. At this stage of the project, the route selection problem is the major objective, while the docking phase of the auto piloted tractors is for the moment neglected, but it will be the next step of the research. The paper is organized as follows: Section 2 contains the problem formulation and a specific example regarding our airport test case; In Section 3 is reported the proposed algorithm to solve the shortest path optimization problem; Section 5 describes the code implementation and shows the results obtained for the selected test case; pertinent conclusion are reported in the closing section.
12
Embed
A route selection problem applied to auto piloted aircraft ... fileA route selection problem applied to auto piloted aircraft pushback tractor M. Cassaro and G. Sirigu Politecnico
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A route selection problem applied to auto
piloted aircraft pushback tractor
M. Cassaro and G. Sirigu
Politecnico di Torino, Turin, Italy
Abstract. Airplanes taxiing on taxiways in airports burn a large amount of fuel, emit tons of CO2, and are
very noisy. The aviation industry demands alternative means to tow airplanes from gate to take-off with
engines stopped (dispatch towing). Pushback tractors are the main candidate to accomplish this mission.
The major issues that must be solved to make this solution applicable are mainly two: structural and
regulatory limitations. From a structural point of view, new technologies are already under investigation to
optimize the towbarless tractors joint and drastically reduce the fatigue loads on the nose landing gear
(NLG). From a regulation point of view, to guarantee safety during towing an automatic control system for
the tractors, capable to control taxiing speed and route inside the airport area, will be the solution. In this
paper, we present a software solution for a route selection problem in a discretized airport environment. The
algorithm, implemented using Hopfield neural network, is able to compute the shortest allowed path in terms
of checkpoints. These, once passed to the tractor autopilot, should be able to guide it from the airport tractor
parking zone to the selected parked aircraft (phase 1), perform taxiing (phase 2) and going back from the
runway threshold to the parking area (phase 3). The phases are in reverse order when landing occurs.
1 Introduction A quite significant revolution is currently undergoing in the aircraft ground operations system. The taxi
phase has always been critical for several reasons such as fuel consumption, noise, pollution and foreign
object damage (FOD). All these issues could be resolved all in once by having the possibility of keeping the
engines off during ground maneuvers. Therefore, a new way of thinking at the pushback tractor is raising in
the aeronautical industries. Instead of using them only to exit the parking lot, allow the aircraft towing from
gate to take-off position and from landing arrival point to the gate. This new usage of the pushback tractor
introduces some problematic from structural and regulatory point of view. Structurally NLG are not designed
to be towed for long distances, which fact can lead to fatigue failure due to transversal loads. To this
purpose, recently patented towbarless system has been designed, built and tested by an industry consortium
to avoid solid link between the tractor and the landing gear structure. The other critical issue is about the
aviation regulation, since aircraft not having a Pilot in Control (PIC), when towed by the tractor driver, face
safety, responsibility and regulatory limitations. Our research team is working on this last aspect by
designing a novel concept of airport ground operation system [1]. It consists in a semi-autonomous system,
activated by the control tower, in which autonomous auto piloted tractors are capable of accomplish towing
missions between points selected by the tower operators. At this stage of the project, the route selection
problem is the major objective, while the docking phase of the auto piloted tractors is for the moment
neglected, but it will be the next step of the research. The paper is organized as follows: Section 2 contains
the problem formulation and a specific example regarding our airport test case; In Section 3 is reported the
proposed algorithm to solve the shortest path optimization problem; Section 5 describes the code
implementation and shows the results obtained for the selected test case; pertinent conclusion are reported in
the closing section.
2 Problem Formulation A single mission of the tractor could be discretize in three different phases that can be execute in opposite
order depending on whether the aircraft has to take-off or had just landed. Phase 1: going from the airport
tractor parking zone to the selected parked aircraft; phase 2: perform taxiing; phase 3: going back from the
runway threshold to the tractor parking area. In each of these phases, the primary problem is to find the
shortest path between two points of the airport. As well known, in the aircraft parking area the ground
vehicles are not allowed everywhere, and some predefined routes are one-way only while some others are
not. Therefore, the straight-line trajectory is obviously not a solution of the problem. Fig. 1, which represents
our test case, shows the actual taxiways and ground vehicle allowed areas in the Caselle airport of Turin. A
constrained route selection problem has to be solved to play the role of path planner every time that a tractor
action is required.
Figure 1 Turin Airport "Caselle", satellite picture with directions for taxi and ground movement.
For the application proposed, it was thought to discretize the airport runways, taxiways and ground vehicle
routes with checkpoints. Figure 2 represents the Caselle airport discretization, where the tractors are allowed
to operate. Then, given as external input (supposedly by the airport tower operator) the tractor identification
number and the aircraft parking lot number, the algorithm should be able to select the shortest allowed path
to optimize the tractor mission.
3 Proposed Algorithm The proposed algorithm consists of two parts: airport data loading and domain discretization, and shortest
path computation by means of Hopfield neural network. Since the pushback tractor mission consists itself of
three different phases, as already explained in the previous paragraph, the first and second part of the
algorithm are compiled only once, whereas the computation of the shortest path is repeated three times, once