Precision Agriculture Applied to Harvesting Operations through the Exploitation of Numerical Simulation

Area of Science:

• Agricultural Engineering
• Robotics and Automation
• Precision Agriculture

Background:

• Optimizing combine harvester operations is crucial for minimizing harvest losses and operational downtime.
• Precision agriculture demands advanced control systems for efficient and accurate harvesting processes.

Purpose of the Study:

• To propose a comprehensive control framework for combine harvesters using a two-layer algorithm.
• To develop a simulation environment for analyzing harvest loss, time, and fuel consumption.
• To evaluate the performance of a vision-guided Stanley Lateral Controller for path tracking.

Main Methods:

• Developed a two-layer control algorithm: path-planning for harvesting techniques and path-tracking using a vision-guided Stanley Lateral Controller.
• Utilized IPG-CarMaker software to create challenging driving scenarios for emulating wheat harvesting.
• Implemented a Driver-in-the-loop (DIL) framework for comparing autonomous and human driving.

Main Results:

• The controller effectively followed reference trajectories in regular field conditions, achieving zero harvest waste and minimal downtime.
• Harsh road irregularities necessitated re-planning of the reference trajectory, suggesting alternative harvesting methods or header overlap.
• Quantitative and qualitative comparisons of harvesting techniques and the relationship between terrain irregularities and required overlap were presented.

Conclusions:

• The developed control framework demonstrates effectiveness in precision agriculture harvesting, particularly in ideal conditions.
• Adaptations, such as trajectory re-planning or header overlap, are necessary for navigating uneven terrain to ensure complete crop harvesting.
• The proposed DIL framework offers a valuable methodology for evaluating autonomous driving systems in agricultural contexts.