Multi-Axis Force Torque Sensor Calibration for Axis Decoupling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing multi-axis force/torque sensors lack independent linearity between axes, leading to limited measurable data range and accuracy issues in robotic control due to signal correlation among axes.

Innovation Solution

A method and apparatus for calibrating multi-axis force/torque sensors using matrix modification and optimization through an input/output matrix equation to ensure axis independence, involving steps of generating linear functions, calculating correlation equations, and integrating calibration values to enhance linearity and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional force/torque sensors with three-dimensional structures are used, then measurement capability is achieved, but the sensor becomes large and expensive with complex manufacturing

Engineering Contradiction:
Improveforce/torque measurement capabilityVSAvoidthree-dimensional structure complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical three-dimensional sensor structures with a simplified two-dimensional PCB-based sensor design. Instead of using traditional strain gages attached to three-dimensional structures, the invention uses printed circuit board traces as force-sensitive elements, eliminating the need for complex mechanical assemblies while maintaining measurement capability.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent employs thin PCB substrates as the sensor structure, replacing bulky three-dimensional mechanical structures. The PCB traces are printed on flexible or rigid thin substrates, creating a planar sensor that maintains measurement functionality while dramatically reducing structural complexity and size.

Inventive Principle:
Principle #30Flexible shells and thin films

2Adaptability or versatility

If multi-axis force/torque sensors are used to measure forces in multiple directions, then comprehensive measurement is achieved, but signal correlation between axes reduces measurement accuracy

Engineering Contradiction:
Improvemulti-axis measurement capabilityVSAvoidaxis independence and linearity
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent applies matrix transformation to change the mathematical parameters of the sensor output. By multiplying the raw sensor signals by a pre-calibrated transformation matrix, the system converts correlated multi-axis measurements into independent axis measurements, effectively decoupling the signal correlations while maintaining multi-axis measurement capability.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent implements a calibration process where the transformation matrix is determined through feedback from known test forces. The matrix is calculated based on actual sensor responses to calibrated test inputs, and this feedback information is used to optimize the transformation that achieves maximum axis independence.

Inventive Principle:
Principle #23Feedback

3Ease of manufacture

If sensor calibration is performed using conventional methods, then basic functionality is achieved, but linearity and independence of each axis are not ensured

Engineering Contradiction:
Improvecalibration process simplicityVSAvoidaxis linearity and independence
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent performs preliminary calibration to determine the transformation matrix before actual use. By pre-calculating the optimal transformation matrix based on manufacturer test data and known force applications, the system prepares the calibration parameters in advance, ensuring accurate axis independence without requiring complex real-time calibration procedures.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses test force data from manufacturer calibration as a reference model. By copying and analyzing the known test force relationships, the system derives the transformation matrix that replicates ideal axis independence, using the manufacturer's calibration data as a template for optimizing the sensor output.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP4671720A1Method for calibrating multi-axis force torque sensor, and apparatus therefor
Publication Date: 2025.12.31 KYUNGPOOK NAT UNIV IND ACADEMIC COOP FOUND
  • EP4671720A1 patent drawingFigure 1~2
  • EP4671720A1 patent drawingFigure 3
  • EP4671720A1 patent drawingFigure 4

AI summary

The present invention relates to a method for calibrating a sensor of a multi-axis force/torque sensor which has a maximum of n axes, measures a force in a vertical direction on the n/2 axes, and measures a torque in a horizontal direction on the n/2 axes, and more particularly, to a method for calibrating a sensor of the multi-axis force/torque sensor which comprises: a first step of preparing test reference data of the multi-axis force/torque sensor; a second step of obtaining sensor output values measured on the n axes provided in the multi-axis force/torque sensor by using the test reference data prepared in the first step as an input value; a third step of calculating a linear relationship between an input value and a sensor output value on each axis for a force and a torque by using the sensor output values obtained in the second step; and a fourth step of calculating an input value corresponding to the sensor output value obtained in the second step based on the linear relationship calculated in the third step, to calculate a correlation equation for a correlation between each axis.