Evaluating the effect of tire parameters on required drawbar pull energy model using adaptive neuro-fuzzy inference system

Determination of the required energy for drawbar pull of agricultural tractors plays a significant role in the characterization of the quality of tractors during different operations. Assessment of the effect of some tire parameters on drawbar pull energy was performed utilizing a single-wheel tester in a soil bin facility. To this aim, the potential of a global searching soft computing approach (i.e. adaptive neuro-fuzzy inference system) with various membership functions was evaluated. The tire parameters of velocity at three levels of 0.8, 1 and 1.2 m/s, wheel load at three levels of 2, 3 and 4 kN and slippage at three levels of 8, 12 and 15% were applied to single-wheel tester while four installed load cells were responsible for the measurement of drawbar pull. It was concluded that drawbar pull energy is a direct function of wheel load, velocity and slippage. Hence, the greatest value of 1.056 kJ corresponded to the wheel load of kN, slippage of 15% and velocity of 1.2 m/s. The outperforming model yielded mean square error and coefficient of determination values of 0.00236 and 0.995, respectively.