A New Perspective on Low-Cost MEMS-Based AHRS Determination.

Area of Science:

• Robotics and Control Systems
• Sensor Fusion and Navigation

Background:

• Attitude and Heading Reference Systems (AHRS) are crucial for determining a rigid body's 3D orientation.
• Accurate AHRS estimation relies on integrating data from gyroscopes, accelerometers, and magnetometers.
• Existing iterative Kalman filtering methods for AHRS require precise initialization.

Purpose of the Study:

• To develop and validate a novel initialization method for AHRS.
• To enhance the accuracy of initial attitude estimates in navigation systems.
• To reduce the computational complexity of AHRS estimation.

Main Methods:

• Integration of low-cost MEMS-based gyroscope, accelerometer, and magnetometer data.
• Implementation of a complementary adaptive Kalman filter for AHRS estimation.
• Development of a new iterative initialization process for attitude estimation.

Main Results:

• Achieved a significant 45% improvement in initial orientation accuracy compared to existing techniques.
• Reduced the computational process by 96% through the proposed initialization method.
• Validated attitude estimates against a high-end reference orientation system.

Conclusions:

• The proposed AHRS initialization method enhances initial attitude accuracy.
• The novel approach substantially decreases computational requirements for AHRS.
• This method offers a more efficient and accurate solution for navigation systems.