Potential assessment of neuro-fuzzy strategy in prognostication of draft parameters of primary tillage implement

This study investigates potential of neuro-fuzzy strategy in prognostication of draft parameters of primary tillage implement. To this aim, computer simulation environment of adaptive neuro-fuzzy inference system (ANFIS) was employed to simulate field data of tillage operations with moldboard plow implement. The field trials were conducted at three levels of forward speed (2, 4 and 6 km/h) and three levels of plowing depth (10, 20 and 30 cm). The plowing depth and forward speed were marked as independent input variables and the draft parameters (draft force and specific draft force) were labeled as dependent output variables in the ANFIS simulation environment. The ANFIS results were compared to those obtained by the equation standardized by American Society of Agricultural and Biological Engineers (ASABE) based on statistical descriptor parameters. Results revealed that the outperforming ANFIS model with acceptable statistical descriptor parameters was more accurate than the ASABE model for prognostication of the draft parameters. The ANFIS modeling results presented that simultaneous increment of forward speed and plowing depth resulted in nonlinear increment of draft force from the lowest bound (<4 kN) to the highest bound (>20 kN). Meanwhile, forward speed increment along with plowing depth decrement resulted in nonlinear increment of specific draft force from the lowest bound (<32 kN/m2) to the highest bound (>120 kN/m2). Furthermore, interaction of forward speed and plowing depth on draft force was congruent. However, it was incongruent in case of specific draft force. According to potential of the ANFIS model assessed in this study, the proposed model can be served as an efficient alternative modeling tool for direct prognostication of the draft parameters of an implement during tillage operations associated with concurrent changes of forward speed and plowing depth.