Design and performance analysis of indoor calibration device for the force-measuring system of the tractor three-point hitch

Abstract

The real-time monitoring of the load in farming by the sensor installed on the tractor's three-point hitch can effectively improve the farming efficiency and force-position combined control, reduce the compaction risk of the wheel on the soil and reduce the fuel consumption in farming process. However, the measurement and quantification of the loads on the three-point hitch have some problems remaining unresolved: testing the accuracy and reliability of a load measuring system is hard when the tractor works in a field, the mathematical model of spatial forces usually lacks a practical and effective validation, and the calibration process of the measurement system is inconvenient and incomplete while easily causing a low accuracy. Specifically, this paper builds a new spatial-force mathematical model based on the geometry of a three-point hitch.To eliminate the discrepancy of the geometric model with the actual structure and to refine the mathematical model, a calibration process is conducted by developing a calibration bench, which is equipped with a data acquisition system and a multi-parameter monitoring interface. The three-point hitch installed on this calibration bench is subject to steady-state loading. The loading force, angle of the lower drawbar, and three-component forces (three shaft pin sensors' forces) of the three-point hitch are well measured. With applying for the measured data to calibrate the theoretical mathematic model eventually derives the resultant force from all the three-component forces, a dynamical loading bench was developed to test the calculated resultant force for the three-point hitch during the sinusoidal and randomly variant dynamical loadings tests. A hitch force measurement system is also developed to collect real-time data and calculate the resultant force of measured three-component forces through the calibrated mathematical model. The results of the dynamical loading tests show that the average relative error MRE = 1.09% with an average force measurement time delay being Δt = 0.5 s, the root mean square error RMSE = 59.3 N, and the coefficient of determination R 2 = 0.9903. As observed, the shape and the trend of the generated resultant force curve are basically the dynamical loading force. The dynamical loading test proves the high efficacy and reliability of the proposed indoor calibration method for calculating the load based on the three-component forces as measured on the three-point hitch. Besides, the preliminary study of the proposed method on the hitch load provides great potential to improve the indoor sixcomponent measurement and quantification of both the force and momentum acting on the three-point hitch.