SIMULATION EXPERIMENTS WITH AN EMPTY RAILWAY WAGON DISTRIBUTION MODEL

The paper describes an empty railway wagon distribution model and some simulation experiments performed with it. The model gives a fairly detailed description of the flow of empty and loaded wagons between different regions, the generation of demand for wagons and the costs connected with transportation and shortage. The purpose of the model is to make experiments possible, mainly to compare different redistribution strategies for the available "empties", but also to test the effect of changes in the "environment", e.g. Changes in the demand pattern. Great care has been used in programming the model so as to get low variance estimates in comparisons of different strategies of environments. This is done by an elaborate "synchronization" of the generated (pseudo) random numbers. (using among other devices an inverted random number generator, generating the random numbers in the reverse order). This synchronization proved quite successful, reducing variances by a factor of at least 5. Antithetic sampling has also been used with success to reduce variances in single runs cutting simulation time by a factor of about 2. The simulations are started near steady state, in that the runs start from an analytically computed average state. Another feature worth mentioning is that other values than the "true" ones for the parameters are tried in the redistribution strategies.

  • Corporate Authors:

    Royal Institute of Technology, Sweden

    Valhallavagen 79
    Stockholm,   Sweden  100 44
  • Authors:
    • Lindberg, P O
    • Svanberg, K
  • Publication Date: 1976

Media Info

  • Features: References;
  • Pagination: 12 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00183287
  • Record Type: Publication
  • Source Agency: Transport and Road Research Laboratory (TRRL)
  • Report/Paper Numbers: TRITA-MAT-1976-15 Monograph
  • Files: ITRD, TRIS
  • Created Date: Nov 14 1978 12:00AM