Rahim Amoozadeh - MA., Department of Agriculture, Islamic Azad University, Shooshtar Branch, Shooshtar, Iran
Mohammad Javad Sheykh Davoudi - Associate Professor, Department of Agriculture, Shahid Chamran University of Ahvaz, Ahvaz, Iran
Heidar Ali Ghasemnejad - Department of Agriculture, Islamic Azad University, Shooshtar Branch, Shooshtar, Iran

Predicting the maintenance and management costs and replacement age of tractors in agricultural mechanized units, is important from several points. So, doing a timely agricultural operations, more accurate measure of the amount of income including the cost of these items , determiningthe useful life of old tractors , replacement age, cost of the process changes and the possibility of examining the undesirable causes of increase in costs are considered. The main data collected for this study include: the amount of the annual maintenance costs which are obtained of three majormaintenance costs (spare parts, maintenance fee, and lubricants). The amount of annual work of tractors is obtained, too. Separating the cost items shows that the cost of spare parts is highest among the total maintenance costs. The final model obtained in this study compared with someother sources indicates that the cumulative amount of these expenses as a percentage of the original price of the model resource estimate is lower than other models. The main reason which has been reported is the difference of maintenance costs (especially in the early years of theoperation of tractors) with an initial price of tractors in comparison with other sources. Finally,the economic life of the machines (the optimal replacement age) was estimated for each model of tractor. Results indicate that among the three models, estimated useful life of the John Deere 3140 Tractor and TM155-Newhland has been minimum and maximum, respectively.

Harvester, Replacement age, Mathematical model, Neural network