Algorithm For Scheduling Runway Operations Under Constrained Position Shifting

Algorithm for Scheduling Runway Operations
Under Constrained Position Shifting

A. Shringi, M. Tiwari, P. Jangir*
School Of Computer Science And Engineering
Vellore Institute Of Technology, Chennai
pankaj.jangir2011@vit.ac.in
1. Introduction
The air transportation system is a tightly constrained system that is operating at (or close to) capacity at most
major airports. In 2005, terminal-area congestion accounted for only 13% of all delays at the 35 busiest
airports; that number had risen to 17% in 2008 and is currently at 21% over the first nine months of 2009.
The increasing delays, coupled with the expected increase in the demand for air transportation in the future,
have motivated several initiatives for the enhancement of terminal-area capacities (Arkind 2004, Boehme
1994). The runway system has been identified as the primary bottleneck in airport capacity, due to various
operational constraints on runway operations. Consequently, even small enhancements to runway throughput
can have a significant impact on system wide delays. The terminal area is a dynamic and uncertain
environment, with constant updates to aircraft states being obtained from surveillance systems and airline
reports (Atkins and Brinton 2002). The dynamic nature of the terminal area necessitates the development of
scheduling algorithms that are computationally efficient and therefore amenable to replanning when new
events occur, such as when a new aircraft enters the centre boundary or when data updates flexibility in
reordering aircraft, however large deviations from a nominal schedule may be unacceptable to airlines from a
fairness standpoint. This observation led to the constrained position shifting (CPS) approach for scheduling
aircraft, which stipulates that an aircraft may be moved up to a specified maximum number of positions from
its FCFS order. For example, if the maximum position shift (MPS) allowed was 2, an aircraft that is in the 8th
position in the FCFS sequence can be placed at the 6th, 7th, 8th, 9th or 10th position in the new sequence.
Several researchers have used CPS to model fairness and have worked toward developing fast solution
techniques for scheduling within the CPS framework
This report presents an algorithm for efficient runway scheduling on a single runway with CPS constraints,
while accounting for various operational considerations (including time-window restrictions and precedence
constraints. The core problem we consider is that of maximizing runway throughput (the landing time of the
last of a given set of aircraft) for arrivals-only or departures-only operations. Consequently, most practical
implementations resort to heuristic or approximate approaches that produce good solutions in a short-time.
The difficulty in solving these scheduling models arises primarily because the solution space allows for the
optimal sequence to deviate arbitrarily from the first-come-first-served (FCFS) sequence.
Keywords
FCFS, Constrained Position Shifting, Runway scheduling, Operations Research
2. Problem Definition
The runway scheduling problem is to find a sequence and corresponding arrival/departure times that optimize
Some objective of the schedule (for example, minimize the spending time in air or minimize a weighted
average of aircraft delay), subject to the following constraints.
1. Position Shift Constraint
As the airlines are major stakeholders in the air transportation system, it is important that an increase in
efficiency is not achieved at the expense of an equitable allocation of resources. This could happen if an
aircraft that would have had an early arrival or departure in the FCFS sequence is rescheduled to operate last,
thereby incurring a disproportionate amount of delay. CPS ensures some degree of fairness because it does
not allow the final sequence to deviate significantly from the FCFS order. The maximum number of position
shifts allowed is denoted by k, and the resultant scenario is referred to as a k-CPS scenario.
2. Minimum spacing requirements
An aircraft operating on a runway faces the risk of instability if it interacts with the wake-vortex of an aircraft
landing or taking off before it. To prevent this, the Federal Aviation Administration (FAA) mandates
minimum spacing requirements under instrument approach conditions (IAC) between aircraft operations on a
runway, which depend on the on the maximum take-off weight capacity of the aircraft. Although these
spacing requirements are specified in terms of distance, They can be converted to time requirements.
Representative values for these separations for three weight classes - small, large, and heavy
3. Time-window constraints
Limits on the levels of delay that can be incurred by an aircraft due to downstream traffic flow management
initiatives or constraints on possible manoeuvres that can be performed by the aircraft restrict the times at
which an aircraft can reach a runway. These constraints could possibly result in a set of disjoint time intervals
in which an aircraft can arrive/depart. For simplicity of notation, the case of a continuous-time interval
defined by an earliest and latest time, but the approach is applicable to disjoint intervals as well.
4. Precedence constraints
Precedence constraints are pairwise requirements on aircraft that stipulate whether one aircraft must land
before another. Sources of such constraints are the airlines themselves, which have precedence constraints
due to banking operations or priority flights. In addition, arrivals on the same jet route are constrained to not
overtake each other. Precedence constraints can also represent the restricted freedom available to taxiing
departures that are not allowed to overtake each other.
3. The CPS Network

The network is a directed acyclic graph in which every feasible CPS sequence is represented by a path in the
network; the scheduling problem is then solved using dynamic programming. For simplicity, an assumption is
made that the aircraft are labelled 1 2.n according to their position in the FCFS sequence. The network
consists of n stages 1n, where each stage corresponds to an aircraft position in the final sequence. A node
in stage p of the network represents a subsequence of aircraft of length min (2k + 1, p). For example, for n =
6 and k = 1, the nodes in stages 3 represent all possible sequences of length 2k + 1 = 3 ending at that stage.
Stage 2 contains a node for every possible aircraft sequence of length 2 ending at position 2, whereas stage 1
contains a node for every possible sequence of length 1 starting at position 1. This network, shown in Figure,
is obtained by finding all sequence combinations of possible aircraft assignments to each position in the
sequence. The last aircraft in a nodes sequence as the final aircraft of that node. Two nodesa source and a
sink are introduced that represent the beginning and end of the sequencing process, respectively. We add
arcs from the source to each node in stage1 and from each node in stage n to the sink. An arc i..j is drawn
from a node in stage p to one in stage p + 1 if the aircraft subsequence of node j can follow that of node i, i.e.,
the first min 2k p aircraft of node js subsequence are the same as the last min (2k,p) aircraft of node is
subsequence. For example, a sequence (123) in stage 3 can be followed by the sequences (234) or
(235) in stage 4. This results in a network where every directed path from a node in stage 1 to one in stage
n represents a possible k-CPS sequence. For example, the path 2 21 213 134 346
465 represents the sequence 213465.
CPS network for n =6, k =1 generated from possible aircraft assignments
4. Complexity
The complexity of the algorithm for finding the minimum make span for n aircraft and maximum position
shift of k is O (n (2k + 1) ^ (2k+2)). The nodes in each stage of the network are generated by all combinations
of length 2k + 1, where each position in the sequence has at most 2k + 1 possible aircraft. The number of
nodes in each stage is therefore O ((2k + 1) ^ (2k+1)), because there are n stages, the total number of nodes in
the network is O (n (2k + 1) ^ (2k+1)). Each node can have at most 2k + 1 predecessors because the sequence
of a node differs from the sequence of its predecessor only in the first and last position. So the number of arcs
is O (n (2k + 1) ^ (2k+2)). Pruning the network requires looking at each arc at most twiceonce during the
forward pass and once during the backward pass. The dynamic programming recursion examines each arc at
most once, so total complexity is equal to the number of arcs in the network, which is O (n (2k + 1) ^
(2k+2)).In the presence of precedence constraints, each node would require O(k) work because we would
need to check if precedence is violated between the final aircraft in the node and each of the other aircraft in
the subsequence of the node. The complexity for pre-processing the entire network would thus be the number
of nodes times, which is the same as the running time of the pruning and dynamic programming recursion.
Although the complexity is exponential in k, it is of little consequence because k is typically small.
Therefore, the complexity of the algorithm is essentially linear in n.
5. Future research direction

Although the proposed framework can solve a versions of the runway-scheduling problem considered so far,
it is not immediately applicable to complex objectives such as the minimization of the sum of arbitrary
Aircraft-dependent costs or to situations in which the separation requirements do not satisfy the triangle
inequality (for example, for mixed arrival-departure operations on a runway). In this section, we present
discrete-time models for solving these extensions. We assume that all input data (such as separations, time
windows, etc.) are integer multiples of some time period. For instance, an interval of about two or five
seconds is appropriate in the context of arrival scheduling given that the separations are of the order of a
minute and it is difficult to control aircraft positions and landing times to a very high degree of accuracy. We
then create a copy of the CPS network for each time period of interest and solve the problems using dynamic
programming on this discrete-time CPS network. Because the number of nodes and arcs in the network now
depends on the number of time periods, the algorithms presented in this section have pseudo polynomial
complexity (polynomial in the number of time periods).
6. Conclusion
The model represents a unified framework for runway scheduling under constrained position shifting and has
demonstrated that the problems of enhancing throughput, decreasing delay, and ensuring fairness can be
effectively modelled and solved in polynomial time (linearly in the number of aircraft) while accounting for
most operational constraints. An extension of the framework to include more general cost functions (by using
discrete-time models) and to mixed arrival and departure operations, including the merging of multiple
departure queues can also be developed. The algorithms can be easily implemented, and a prototype
implementation for arrival scheduling demonstrates that the run times are sufficiently small to enable
real-time deployment.
7. References
7.1 Anagnostakis, I., J.-P. Clarke, D. Bhme, U. Vlckers. 2001. Runway operations planning and control:
Sequencing and scheduling. J. Aircraft 38(6) 988996.
7.2 Lee, H. 2008. Trade-off evaluation of scheduling algorithms for terminal-area air traffic control. Masters
thesis, Massachusetts Institute of Technology, Cambridge, MA.
8. Attachments
-input.txt
-filerunway.py
-output.txt

Algorithm For Scheduling Runway Operations Under Constrained Position Shifting

Uploaded by

Copyright:

Available Formats

Algorithm For Scheduling Runway Operations Under Constrained Position Shifting

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Algorithm For Scheduling Runway Operations Under Constrained Position Shifting

Uploaded by

Copyright:

Available Formats

Algorithm for Scheduling Runway Operations

Under Constrained Position Shifting

3. The CPS Network

CPS network for n =6, k =1 generated from possible aircraft assignments

5. Future research direction

You might also like