Investigating the effect of the tractor driving system type on soil compaction using different methods of ANN, ANFIS and step wise regression

Abstract

Farm tractors have different driving systems, including four-wheel drive and rear-wheel drive which are common systems employed to conduct agricultural operations all over the world. The type of driving system influences tractor wheel slip which is one of the most effective factors in soil compaction. Hence a series of experiments were conducted using Goldoni 240 tractor to investigate the effects of tractor driving systems, including four-wheel drive (4WD), rear-wheel drive (RWD), and front-wheel drive (FWD) at different travel speeds, soil moistures, number of passes (1, 5, and 9), and soil textures (clay loam, loam, and sandy loam) at the depths of 10, 20, 30, and 40 cm. Increasing tractor wheel slip showed a significant effect on increasing soil compaction. The lowest tractor wheel slip was occurred using the 4WD system, and by increasing travel speed in this system, soil compaction decreased. Increasing tractor speed in the FWD system increased tractor wheel slip and soil compaction. In addition, increasing soil moisture content resulted in an increase in soil compaction, and this event was intense in fine soils like clay loam. It was found that adaptive neuro-fuzzy inference system (ANFIS) higher potential to predict the effect of multiple input variables on soil compaction (R2 =0.99) than regression method. According to the standard coefficients of regression models, depth, type of driving system, number of passes, moisture content, texture, speed, and inflation pressure were the factors significantly influencing soil bulk density, respectively.