Prediction of Tractor's Economic Life Using Genetic Algorithm

Replacement of a tractor is an influential factor affecting timely farm operation. The accurate prediction of repair and maintenance costs is essential to selection of a replacement model. This study used empirical data for 60 two-wheel drive (2WD) tractors from Astan Ghods-e Razavi agroindustry. The types and numbers of tractors used were the Massey Ferguson 285 (17), Fiat (10), John Deere 3140 (28) and John Deere 4450 (5). Regression analysis showed that the quadratic model was suited for predicting repair and maintenance costs. The four types of tractors recorded 17850, 18380, 27000 and 2740 cumulative hours, respectively, of use to replacement life as predicted using the genetic algorithm.