A dynamic model for GPS based attitude determination and testing using a serial robotic manipulator

Area of Science:

• Robotics
• Geomatics Engineering
• Computational Kinematics

Background:

• Accurate attitude estimation is crucial for robotic arm control and navigation.
• Dynamic environments pose challenges for real-time kinematic tracking.
• Existing methods may lack precision in complex, swinging motions.

Purpose of the Study:

• To develop and validate a computational algorithm for precise robotic arm attitude determination.
• To enable accurate 3-axis attitude estimation for manipulators moving along predetermined paths.
• To integrate static and dynamic modes for comprehensive attitude assessment.

Main Methods:

• Utilizing three Global Positioning System (GPS) L1 receivers for simultaneous measurements.
• Processing GPS data through conversion to RINEX format and baseline vector determination using the Least-Squares Ambiguity Decorrelation (LAMBDA) method.
• Employing Least-Squares Attitude Determination (LSAD) in static mode and an extended Kalman filter in dynamic mode for attitude calculation.

Main Results:

• Successful application of the algorithm to a Mitsubishi RV-M1 robotic arm.
• Accurate attitude estimates generated for the robotic arm's dynamic movements.
• Validation of results through independent evaluation of Euler angles from robotic arm postures.

Conclusions:

• The developed algorithm provides accurate and reliable attitude estimation for robotic arms.
• The integration of static and dynamic modes enhances the precision of attitude determination.
• This approach offers a robust solution for real-time kinematic tracking in robotics.