A Loosely Coupled Extended Kalman Filter Algorithm for Agricultural Scene-Based Multi-Sensor Fusion.

Area of Science:

• Robotics and Automation
• Agricultural Technology
• Sensor Fusion

Background:

• The increasing aging population and advancements in modern agriculture necessitate the development of autonomous agricultural robots for large-scale production.
• Current agricultural robot navigation systems are susceptible to failures caused by external noise and environmental factors, hindering reliable operation.

Purpose of the Study:

• To propose and validate a novel multi-sensor fusion method for agricultural robots to enhance navigation system robustness and accuracy.
• To mitigate the impact of external environmental interference on the navigation system's performance.

Main Methods:

• Development of an agricultural scene-based multi-sensor fusion method utilizing a loosely coupled extended Kalman filter algorithm.
• Integration and fusion of data from multiple sensors: Inertial Measurement Unit (IMU), Robot Odometry (ODOM), Global Navigation and Positioning System (GPS), and Visual Inertial Odometry (VIO).
• Utilized visualization tools for simulating and analyzing robot trajectory and error.

Main Results:

• The proposed multi-sensor fusion algorithm demonstrated high accuracy and robustness in experimental evaluations, even during sensor failures.
• Comparative analysis showed superior accuracy and robustness of the proposed algorithm on an agricultural dataset compared to existing methods.
• Successful simulation and analysis of robot trajectory and error using visualization tools confirmed the algorithm's effectiveness.

Conclusions:

• The developed agricultural scene-based multi-sensor fusion method effectively reduces external environmental interference, enhancing navigation system reliability.
• The proposed method offers a robust and accurate solution for autonomous navigation in agricultural robots, addressing critical challenges in the field.
• This advancement is crucial for the future trend of large-scale agricultural production using autonomous robotic systems.