Robot Wrench Estimation Using Partial Sensing and Dynamic Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotics systems face challenges in accurately estimating wrench forces and torques without the need for expensive 6 degrees of freedom (DOF) force/torque sensors, as the precision of dynamic models depends heavily on predicting friction parameters that may not be known or measurable with desired precision.
Innovation Solution
A method that measures at least one component of the wrench using built-in sensors and estimates the remaining components based on a dynamic model of the robot, utilizing techniques such as maximum-likelihood estimation, Kalman filters, and maximum-aposteriori estimation to improve precision at reduced costs by minimizing the number of sensors required.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a 6 DOF force/torque sensor is used to measure all wrench components, then measurement precision is improved, but device cost and complexity increase significantly
Solution Approach 1:
The wrench measurement task is segmented into two parts: (1) direct measurement of selected wrench components using built-in robot sensors, and (2) estimation of remaining components using dynamic model-based algorithms. This segmentation avoids the need for expensive 6 DOF F/T sensors while maintaining measurement precision through hybrid measurement-estimation approach.
Solution Approach 2:
A dynamic model of the robot acts as an intermediary to bridge the gap between limited sensor measurements and complete wrench estimation. The dynamic model uses motor current data and robot kinematics to infer unmeasured wrench components, enabling accurate force/torque estimation without direct sensor measurement of all components.
2Device complexity
If dynamic model-based estimation is used for all wrench components, then device cost is reduced, but measurement precision deteriorates due to friction prediction difficulties
Solution Approach 1:
The solution merges direct sensor measurement with dynamic model estimation in a hybrid approach. Built-in robot sensors directly measure selected wrench components (providing high precision for those components), while the dynamic model estimates remaining components. This combination leverages the strengths of both methods to achieve overall high precision at reduced cost.
Solution Approach 2:
The approach changes the parameter measurement strategy by selectively measuring only certain wrench components (e.g., forces in specific directions) rather than all six components. This parameter selection optimizes the trade-off between measurement precision and device cost, allowing critical parameters to be directly measured while less critical ones are estimated.
3Measurement precision
If friction parameters are accurately predicted in the dynamic model, then wrench estimation precision is improved, but device complexity and calibration requirements increase
Solution Approach 1:
Instead of attempting to accurately model all friction parameters (which would require complex calibration), the approach uses partial action by relying on direct sensor measurements for critical wrench components where precision is most important. The dynamic model with simplified friction assumptions suffices for estimating remaining components, avoiding the complexity of comprehensive friction calibration.
Data Source
AI summary
A method for estimating a wrench acting on a reference point of a robot includes the steps of: a) measuring at least one component, but not all components, of the wrench; and b) estimating non-measured components of the wrench based on a dynamical model of the robot while taking into account the measured components.