Robot Interaction Control With IMU-Based Inertial Force Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Real-time robot control systems face instability due to unaccounted forces from high payload objects and unmodeled structural dynamics, which confound motion and lead to inaccuracies in sensor readings and motion control, especially in high-payload settings.
Innovation Solution
Incorporating an inertial measurement unit (IMU) alongside a force-torque sensor to adjust force values and correct for accelerations from the robot's motion and unmodeled structural dynamics, allowing for more precise interaction control and improved motion fidelity by directly sensing unwanted forces and structural oscillations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex modeling techniques are used to compensate for post-sensor inertia, then some motion accuracy can be achieved, but the system cannot account for all structural dynamics and is not suitable for real-time control with strict timing requirements
Solution Approach 1:
The patent replaces complex mechanical modeling approaches with direct sensor measurement. Instead of using complex mathematical models to estimate and compensate for inertial forces, the system directly measures actual accelerations using an IMU and computes the inertial force contribution through simple multiplication (F=ma), providing accurate real-time compensation without complex modeling
Solution Approach 2:
The system uses the robot's own motion sensors (IMU) to measure and compensate for its own inertial effects. The IMU mounted on the end effector directly measures the accelerations that generate post-sensor inertial forces, and the control system uses this self-measured data to compensate for the effects, making the system self-correcting without external complex modeling
2Ease of operation
If admittance control is used for robot interaction, then robot compliance with the environment is improved, but unaccounted forces from high payload objects cause instability
Solution Approach 1:
The patent implements direct feedback from the IMU acceleration measurements to the admittance control loop. The measured accelerations are used to compute inertial force contributions that are then subtracted from the force sensor readings before being processed by the admittance controller, providing real-time feedback compensation that maintains stability while preserving compliance
Solution Approach 2:
The system converts the harmful effect of unaccounted inertial forces into a beneficial compensation mechanism. By measuring the actual accelerations that cause the inertial forces and using them to compute and subtract the force contributions, the system turns the problem of high payload inertia into an opportunity for precise real-time compensation, improving both stability and control accuracy
3Device complexity
If force-torque sensor alone is used for interaction control, then system simplicity is maintained, but unwanted forces from robot motion and unmodeled structural dynamics cannot be sensed or corrected
Solution Approach 1:
The patent merges the force-torque sensor measurements with IMU acceleration measurements to achieve superior control. By combining the direct force measurements from the FT sensor with the acceleration measurements from the IMU, the system can separate actual interaction forces from inertial forces, providing both the simplicity of direct force sensing and the accuracy of inertial compensation
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the robot's compliance and interaction with the workcell, providing higher fidelity motion control and reducing cycle times by accounting for confounding forces and dynamics, thus enabling more stable and efficient task execution in real-time.
Implementation Method 1
a sensor suite including, but not limited to, an inertial measurement unit (IMU) 110, a force-torque (FT) sensor 108
Data Source
AI summary
Methods, systems, and apparatuses, including computer programs encoded on a computer storage medium, for adjusting the force value received in an interaction control system using an inertial measurement unit to account for post-sensor inertia. In one aspect, the method can include an interaction control system receiving an updated force value representing a force on the last movable component of a robot and an updated acceleration value generated by an inertial measurement unit mounted on the last component of the robot, and adjusting the force value received from the force-torque sensor based on the acceleration value received from the inertial measurement unit to generate an updated command that accounts for the post-sensor inertia.