Prediction of Draft Force and Disturbed Soil Area of a Chisel Tine in Soil Bin Conditions Using Draft Force and Its Comparison with Regression Model

Authors : Kazım ÇARMAN, Tamer MARAKOĞLU, Alper TANER, Ergün ÇITIL

Pages : 56-63

View : 21 | Download : 35

Publication Date : 2021-04-25

Article Type : Research Paper

Abstract :One of our most valuable natural resources is soil. Sustainable agricultural production is achieved with proper soil management. Tillage is considered to be one of the largest operations, as the most energy need in agricultural production occurs in tillage. The main purpose of this study is to investigate the effects of chisel tine on draft force and disturbed soil area and estimate them using artificial neural networks insert ignore into journalissuearticles values(ANN); and multiple linear regression equations insert ignore into journalissuearticles values(MLR);. The experiments were carried out in a closed soil bin filled with clay loam soil at an average moisture content of 13.2% insert ignore into journalissuearticles values(on dry basis);. The draft force and disturbed soil area were evaluated as affected by the share width at two levels insert ignore into journalissuearticles values(60 and 120 mm);, forward speed at four levels insert ignore into journalissuearticles values(0.7, 1, 1.25 and 1.5 ms-1); and working depth at four levels insert ignore into journalissuearticles values(160, 200, 240 and 280 mm); at three replications. The draft force varied from 0.5 to 1.42 kN, depending on the controlled variables, while the disturbed soil area varied from 260 to 865 cm2. Test results show that share width, forward speed and working depth were significant on the draft force and disturbed soil area. Input variables of the ANN models were considered share width, forward speed and working depth. In prediction of required draft force and disturbed soil area respectively, on account of statistical performance criteria, the best ANN model with coefficient of determination of 0.999 and 0.998, root mean square error of 0.010 and 0.016 and mean relative percentage error of 0.960 and 1.673 was better performed than the MLR model.
Keywords : Artificial Neural Networks, Draft Force, Disturbed Soil Area, Chisel Tine