一种高效的六维力传感器校准方法
By combining sparse sampling and an autoencoder model, the problems of nonlinear error and large sample size in the calibration of six-dimensional force sensors are solved, achieving efficient and accurate calibration of six-dimensional force sensors and improving production efficiency and calibration accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Filing Date
- 2026-06-17
- Publication Date
- 2026-07-17
AI Technical Summary
Existing six-dimensional force sensor calibration methods cannot compensate for nonlinear errors, resulting in low measurement accuracy at the range boundary or under complex force conditions. Furthermore, existing nonlinear calibration methods require a large number of samples and long acquisition times, leading to low production efficiency and model overfitting problems.
A sparse sampling strategy is adopted to obtain initial samples. Combined with an autoencoder model and error-driven active sample selection, virtual samples are generated and incrementally acquired. The autoencoder is trained through iterative optimization to achieve efficient calibration.
It effectively reduces the amount of physical sample collection, shortens the calibration cycle, improves production efficiency, achieves high-precision calibration across the entire range, avoids overfitting, and ensures good generalization ability.
Smart Images
Figure CN122409058A_ABST