Development of Path Generation and Algorithm for Autonomous Combine Harvester Using Dual GPS Antenna

Area of Science:

• Agricultural Engineering
• Robotics
• Autonomous Systems

Background:

• Agricultural fields face challenges addressable by autonomous driving technology.
• Tracked combine harvesters, common in East Asia, possess unique steering dynamics compared to wheeled tractors.
• Existing autonomous systems may not be optimized for the specific control characteristics of tracked agricultural machinery.

Purpose of the Study:

• To develop and validate a dual GPS antenna-based autonomous driving system for robot combine harvesters.
• To create and test novel path generation and tracking algorithms tailored for tracked vehicles in agriculture.
• To evaluate the efficiency and accuracy of the autonomous system compared to manual operation.

Main Methods:

• Development of a dual GPS antenna system for precise localization.
• Implementation of an α-turn-type work path generation algorithm.
• Design and application of a path tracking algorithm specifically for tracked combine harvesters.
• Experimental validation using actual combine harvesters in both harvesting and non-harvesting scenarios.

Main Results:

• Autonomous driving achieved low errors: 0.052 m (working) and 0.207 m (turning) without harvesting; 0.038 m (working) and 0.195 m (turning) during harvesting.
• The autonomous system with harvesting work demonstrated a 76.7% increase in efficiency compared to manual driving.
• Reduced non-work area and driving time were observed in the self-driving experiments.

Conclusions:

• The developed dual GPS autonomous driving system and path tracking algorithms are effective for robot combine harvesters.
• The system significantly improves operational efficiency in agricultural harvesting tasks.
• This technology offers a viable solution for enhancing automation and productivity in agriculture.