Design of an Agricultural Robot for Harvesting Melons

摘要

The performance of an agricultural robot has been evaluated through simulation to determine design parameters for a robotic melon harvester. Animated, visual simulation provided a powerful tool to initiate the evaluation of alternative designs. To quantify the many, closely-related design parameters, numerical simulation tools were developed and applied. Simulations using measured cantaloupe locations revealed the effect of design parameters (configuration, number of arms, and actuator speeds) on the average cycle time. Simulation results predicted that a Cartesian robot would perform faster than a cylindrical robot for the melon harvesting task. Activating two arms in tandem was the fastest configuration evaluated. Additional sets of melon locations were stochastically generated from distributions of the field data to determine performance for planting distances between 25 and 125 cm. The fastest cycle time was achieved for an experimental cultural practice that consisted of one plant on each half row in an alternating sequence with 125 cm planting distance. The performance of the robotic melon harvester was found to be highly dependent on the picking time, actuator speeds and planting distance.