Specific draft modeling for combined and simple tillage implements using mathematical, regression and ANN modeling in silty clay loam soil

Abstract

Specific draft is one of the important parameters in the design of tillage tools and estimation of energy consumption in tillage operations, which makes it useful. For this reason, it’s prediction and modeling about different working conditions (depth of plowing and advance velocity) provide the possibility to determine the proper working conditions of tillage implements. In this study, two groups of tillage implements with different geometry including combined tillage implements (combined tiller and chisel packer) and simple tillage implements (moldboard plow, disk plow, chisel plow and offset disk harrow) are used. Also 3 speeds (3, 4.5 and 6 km / h) and 3 different depths (15, 20 and 25 cm) in silty clay loamy soil (47% silt, 22% sand and 31% clay) with 7% moisture content (based on dry) in the form of split twice plots design on the basis of completely randomized block with three replications was considered and the tractor's performance parameter, including specific draft (force per unit width), was measured. Empirical regression models (Linear and nonlinear), and artificial neural network (ANN) modeling were performed on each implement and were compared to other regression and mathematical models of Kheiralla, ASABE and Goryachkin. The type of tillage, depth and speed of advance showed a significant effect on specific draft. The best model was to predict the specific draft nonlinear regression developed in this study. Also, the linear regression model and ANN model obtained in this research after the nonlinear regression model were a suitable model for predicting specific draft in six types of tillage implements. Among the experimental models, ASABE was the perfect model for all simple tillage implements except for disk plow and disk harrow. Also, this model has the potential to be used for compound tillage implements.