Jump to content

Draft:Chinese Postman Problem

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by WalkingRadiance (talk | contribs) at 01:52, 8 October 2022 (I added more details.). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.


The Chinese Postman problem (CPP) relates to a graph G with vertices V and edges E. The graph G should be finite, undirected, and connected. The edges are often assigned a finite real-valued weight that is not negative. The Chinese Postman Problem has been proven to be reducible in polynomial time, in constrast to the Mixed Chinese Postman Problem (MCPP) which is NP complete.[1]

The CPP forms the basis of the field of arc routing compared to more complex cases including the MCPP, the directed chinese postman problem DCPP, the windy postman problem WPP, the rural postman problem RPP, and the hierarchical chinese postman problem HCPP.[2] The CPP has many important real-world applications from school bus routing to mail and postal delivery routing to meter reading planning to police car patrolling, street-sweeping, snow-plowing, salt-spreading, newspaper delivery, and delivering packages.[2]

The CPP requires finding a way to traverse every edge in the graph at least once with minimal cost. To do this, the graph can be augmented until it is Eulerian by duplicating edges between odd degree vertices. The optimal tour will then traverse the augmented Eulerian graph. This can be computed using the Hungarian algorithm.

See Also

Further Reading

  • A. Corberan, Arc Routing:Problems, Methods and Applications, SIAM, 2015.

References

  1. ^ Haouari, Mohumed (March 2002). "Arc Routing: Theory, Solutions and Applications Moshe Dror (editor) Kluwer Academic Publishers, 2000, ISBN 0-792-37898-9". IIE Transactions. 34 (3): 331–332. doi:10.1080/07408170208936917. ISSN 0740-817X.
  2. ^ a b Hrsg., Corberán, Ángel (2015). Arc routing : problems, methods, and applications. SIAM. ISBN 978-1-61197-366-2. OCLC 904346272.{{cite book}}: CS1 maint: multiple names: authors list (link)