Return to search

Autonomous in-flight path planning to replace pure collision avoidance for free flight aircraft using automatic depedent surveillance broadcast

By the year 2020 the number of aircraft will have
increased substantially and will be in �Free Flight�(that
is, ATC will be devolved to the aircraft rather than
being ground based). As an aid to navigation a more
advanced form of collision avoidance will be required.
This thesis proposes a method of collision avoidance
planning using Automatic Dependent Surveillance-Broadcast
(ADS-B) and Dynamic Programming (DP). It in essence
enables Air Traffic Control (ATC) from within the cockpit
for remote or uncontrolled airspace and is a step toward
Free Flight. Free Flight requires quite different
strategies than those used in the present collision
avoidance schemes.
This thesis reviews the approaches to collision
avoidance used in the Air traffic navigation and to
similar problems in other industries. In particular it
considers the extended problem of collision avoidance
within the framework of path planning. This is a key
departure from the approach to aircraft collision
avoidance used in the industry to date. Path planning
reflects the real goal of an aircraft, which is to reach
a particular destination efficiently and safely. Dynamic
Programming is one solution method used in other
industries for the problem of path planning to avoid
collisions with fixed obstacles. The solution proposed herein for the Aircraft case uses Dynamic Programming
applied to the moving obstacle case.
The problem is first simplified by assuming fixed
(static) obstacles for the cost minimisation algorithms.
These fixed obstacles are then moved with time and the
minimisation process is repeated at each time increment.
Although this method works well in most cases, situations
can be constructed where this method fails, allowing a
collision. A modified approach is then used, whereby the
movement of obstacles is included more explicitly (by
modifying the shapes of the obstacles to represent
motion) in the cost minimisation algorithm and a safe
manoeuvre distance for each aircraft is used (by
expanding the object size), to allow space for aircraft
to execute safe evasive manoeuvres in difficult cases.
This modification allows solutions which are complete
(with no known cases of failure � collision situations)
and should be considered as an important extension to the
current Aircraft and Collision Avoidance System (ACAS).
The testing of these solutions is focussed on the
most difficult cases, and includes aircraft movement in
�real space� (that is simulations using real aircraft
dynamics together with dynamic programming algorithms
running in discrete time steps).

Identiferoai:union.ndltd.org:ADTP/216565
Date January 2003
CreatorsHoldsworth, Robert, roberth@gil.com.au
PublisherSwinburne University of Technology.
Source SetsAustraliasian Digital Theses Program
LanguageEnglish
Detected LanguageEnglish
Rightshttp://www.swin.edu.au/), Copyright Robert Holdsworth

Page generated in 0.0018 seconds