COMS 482: unofficial class blog


Lecture 19: Airline Scheduling Problem

Posted in Class Notes by Elliott Back on March 7th, 2005. [Del.icio.us]

Inputs:

  • Set of flights to cover
  • Pairs of flights that can use the same plane
  • Set of k planes

Output:

  • Schedule showing how to use our k planes to cover the flights

Flow graph G:

    vertices = source, sink

    foreach flight i, start ui & dest vi
        1) foreach flight i, an edge (ui, vi) with capacity 1 and lower bound 1. We want to cover each flight.
        2) foreach flight i and j such that j is “reachable” from i, we have an edge (ui, vi) with capacity 1 and lower bound 0. We want to reuse a plane if possible.
        3) foreach flight i, an edge (s, ui) with capacity 1 and lower bound 0, so that a plane can start its day on any flight.
        4) foreach flight j, and edge (vj, t) with capacity 1 and lower bound 0, so that a plane can end its day anywhere.
        5) an edge (s,t) with capacity k and lower bound 0, so that we don’t have to use all the planes.

Airline scheduling with network flow

Algorithm:
    Construct G
    Find circulation

“There exists a feasible circulation iff there exists a valid schedule”

Pf(=>): Assume a feasible circulation, meaning each flight is covered by lowe bound on 1. We can read off the schedule by finding edges with flow 1.

Pf(=<): Assume a valid schedule. The schedule can be used to define a circulation. A valid schedule implies all flights are covered which implies that all lower bounds are satisfied.

This entry was posted on Monday, March 7th, 2005 at 12:06 pm and is tagged with , , , , , , , , , , , , . You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback.

Leave a Reply

Your blog, no spam: the widespread, extraordinary, precious, and unmitigated WP Hashcash.

Please take time to enjoy the archives: May 2005 (1) April 2005 (11) March 2005 (11) February 2005 (15) January 2005 (7)

Fresh, related resources:

Supplied by Google Blog Search
  • There were no results.