Fuzzy C-Regression Clustering Algorithm Based 7-Dof Redundant Manipulators Inverse Dynamics Control
High-precision tracking control of redundant manipulators is essential in robotic applications, but it often requires an exact dynamic model. However, the complex mechanical structure of these manipulators makes it challenging to apply traditional model-based control methods effectively. To address...
Uložené v:
| Vydané v: | Chinese Control and Decision Conference s. 319 - 324 |
|---|---|
| Hlavní autori: | , , , |
| Médium: | Konferenčný príspevok.. |
| Jazyk: | English |
| Vydavateľské údaje: |
IEEE
16.05.2025
|
| Predmet: | |
| ISSN: | 1948-9447 |
| On-line prístup: | Získať plný text |
| Tagy: |
Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
|
| Shrnutí: | High-precision tracking control of redundant manipulators is essential in robotic applications, but it often requires an exact dynamic model. However, the complex mechanical structure of these manipulators makes it challenging to apply traditional model-based control methods effectively. To address this challenge, we propose a data-driven approach using the interval type-2 (IT2) fuzzy c-regression clustering (FCRM) algorithm to develop an inverse dynamics model for redundant manipulators. Utilizing the FCRM algorithm, the nonlinear inverse dynamics input-output data of the robotic arm is subjected to hyper-planeshaped (HPS) clustering, which results in the generation of a T-S submodel and its associated membership function (MF). With the aid of the IT2 fuzzy set, the T-S model captures unmodeled dynamics and uncertainty information of redundant manipulators. Offline data is utilized to learn the inverse dynamics TS submodel and its membership function, enabling real-time control of the manipulator input based on predicted control torque derived from the target trajectory. Simulation results on the Franka Emika Panda (7-DOF redundant manipulator) validate the feasibility of the proposed method. |
|---|---|
| ISSN: | 1948-9447 |
| DOI: | 10.1109/CCDC65474.2025.11090180 |