Simulation modelling of mechanical systems for intra-row weeding in a precision farming approach

Abstract

Aim of study: To test new approaches to perform mechanical weeding inside the row in horticulture and tree fruit fields. The idea is to weed the row by skipping the crop by means of a rotating system instead of a traditional crosswise one. Area of study: North of Italy. Material and methods: Numerical models have been developed to simulate mechanical weeding over time by generating numerical maps to quantify the different kind of worked areas. Main results: Considering the efficiency of weed control on the row, the rotating plant-skipping system with vertical axis (RPSSVA model) with two working tools gives the best performance index (1.1.RWA% = 95.9%). A similar performance can be obtained by the crosswise displacement plant-skipping system, but with very high crosswise translation velocity (with va/vr ratio = 1/5, 1.1.RWA% = 94.5%). With regard to the outwards worked area the RPSS-VA models give the best performances (2.2.%OWAR index from 127.2% up to 282.3%). To reduce the worked area outside the row, the FBTS models give lower index (2.1.OWAR%), while the RPSS-HA works only on the row, but with the lower 1.1.RWA% index among all tested models (55.8%). Research highlights: Rotating systems resulted more efficient than traditional ones, and provide considerations on the use of electric drive power instead of hydraulic one. This study highlights also the need of new approaches in designing lighter working tools. Lastly, the proposed classification of the worked areas could be used as reference standard.