Design and construction of three point hitch device for measuring draft of tillage implement - Data acquisition and post processing analysis

Abstract

This paper presents testing of a new three-point-hitch system for measuring draft of tillage implements of II and III category. The tests were performed in both laboratory and field conditions in order to confirm the reliability of the measuring system and to analyse the measuring method from the aspect of sampling frequency. The results of calibration showed that the determination coefficient R2was almost 1 for linear function. Field testing was performed in 19 passing which were 400 m long, out of which 17 were on the part of the plot that was not tilled and 2 were on the part of the plot where shallow tillage was previously performed with a disc harrow. On the untilled part of the plot, the lowest average value of draft was 17.36 kN measured in the first passing, and the highest draft of 19.94 kN was recorded in the eighth passing. Standard deviation of draft was the biggest in the first passing (3.31 kN) and the results for the rest of the passing showed similarities in this parameter with values ranging from 1.58 kN in the 16thto 2.38 kN in the 15thpassing. As it was expected, some lower values of draft were recorded in the tilled part of the plot (15.33 kN in the first and 16.03 kN in the second passing) in comparison to the untilled part. The estimated draft force at a depth of 0.25 m, working speed of 1.8 m s-1and soil texture was 17.9 kN, which was almost the same as the average value of 17.4 kN. Measuring method was checked based on the frequency of sensor readings and it was done on the simulated samples. Simulation was used to produce the effect of less frequent measuring in order to determine whether the selected frequency (50 Hz) was satisfactory. The simulation process was carried out by reducing the sample size and comparing the draft. The results and analyses of comparison of original and simulated data showed no statistical differences, thus proving the measuring frequency to be more than sufficient for the testing conditions. Joint-time-frequency analysis of the signal in a real time conditions confirmed the results of analysis of 'simulated samples', which clearly showed that spectral densities were at frequencies of up to 5 Hz.