NECSI Resources

 International Conference on Complex Systems (ICCS2007)

Agent Based Model of Livestock Movements

David Miron
CSIRO Livestock Industries

Irina Emelyanova
CSIRO Livestock Industries

Graham Donald
CSIRO Livestock Industries

Graeme Garner

     Full text: PDF
     Last modified: September 27, 2007

The modelling of livestock movements within Australia is of national importance for the purposes of the management and control of exotic disease spread, infrastructure development and the economic forecasting of livestock markets. In this talk an agent based model for the forecasting of livestock movements is presented. This model currently models livestock movements from farm to farm through a saleyard. The decision of farmers to sell or buy cattle is often complex and involves many factors such as climate forecast, commodity prices, the type of farm enterprise, the number of animals available and associated off-shore effects. In this model the farm agentís intelligence is implemented using a fuzzy decision tree that utilised two of these factors. These two factors were the livestock price fetched at the last sale and the number of stock on the farm. On each iteration of the model farms choose either to buy, sell or abstain from the market thus creating an artificial supply and demand. The buyers and sellers then congregated at the saleyard where livestock were auctioned using a second price sealed bid. The price time series output by the model has properties that are similar to those found in real livestock markets.

Maintained by NECSI Webmaster Copyright © 2000-2007 New England Complex Systems Institute. All rights reserved.