Coverage path planning for capacitated harvest vehicles: Coverage path planning for capacitated harvest vehicles for the inner field

By Thijmen Ros

Abstract
When material such as a fertilizer is distributed over a field, or goods are collected as in the case of harvesting. Capacity constraints apply to vehicles like combine harvesters whose grain bins have a limited capacity. Capacitated vehicles are common in agriculture, the limited capacity has to be considered when panning routes for field work. With the rise of fully and partially autonomous agricultural equipment, the path planning process is shifting from the human operators to the machines themselves. When working in a field, ideally the distance driven is minimised so that the work is done as fast as possible and with the least soil compaction. Capacity constraints influence the path planning process as reloading or unloading has to be considered in the routes. In this research, a method was developed with a stationary unload point at the field edge. Per swath or pass of the field, a capacity demand was determined. Between all relevant nodes in the graphical representation of a field, the edge weights were determined based on the distance between these nodes. With this information, optimisations were done to determine the most efficient route over a field meeting full coverage and never exceeding the vehicle’s capacity. A reference route was generated according to the boustrophedon pattern. A field was selected and for three different harvester models and for four different yield rates, optimisations were performed and reference routes generated. In each of the twelve situations, an improvement in the non-working distance was achieved with the optimisation. Relative improvements between 5.22% and 29.87% were observed with use of the optimiser as opposed to the reference routes. The methodology of optimisation in this report relies on known capacity demands, in harvesting the capacity demands are not known exactly beforehand.