Real-Time Formation Planning for Multirobot Cooperation: A Neural Informatics Perspective
Uloženo v:
| Název: | Real-Time Formation Planning for Multirobot Cooperation: A Neural Informatics Perspective |
|---|---|
| Autoři: | Tinglei Wang, Cheng Hua, Yufei Wang, Xinwei Cao, Bolin Liao, Shuai Li |
| Zdroj: | IEEE Transactions on Industrial Electronics. :1-13 |
| Informace o vydavateli: | Institute of Electrical and Electronics Engineers (IEEE), 2025. |
| Rok vydání: | 2025 |
| Témata: | Planning, Multi-robot systems, Mobile robots, Vehicle dynamics, Robot kinematics, Convergence, Noise, Robots, Real-time systems, Neural networks |
| Popis: | Multirobot formation planning has wide applications across various domains. This article proposes an innovative neural-controller-based approach to address the formation planning problem. The proposed noise-tolerant fixed-time zeroing neural network (NT-FTZNN) controller achieves convergence under constant noise, dynamic bounded noise, and even dynamic unbounded noise, demonstrating strong adaptability in complicated scenarios. To the best of our knowledge, this article is the first application of a neural controller with both noise robustness and fixed-time convergence properties to multirobot formation planning tasks. In the presence of various kinds of noises, the proposed method achieves a formation error on the order of 10−7, which significantly outperforms other advanced formation control methods that typically reach only the 10−2 level. Moreover, rigorous theoretical analysis proves that the proposed controller guarantees global stability and fixed-time convergence under various noise conditions. Extensive numerical simulations and physical experiments further validate the superiority of the proposed approach over existing methods, confirming its practical effectiveness in real-world multirobot formation tasks. |
| Druh dokumentu: | Article |
| ISSN: | 1557-9948 0278-0046 |
| DOI: | 10.1109/tie.2025.3579075 |
| Přístupová URL adresa: | https://cris.vtt.fi/en/publications/1840cd51-a218-415e-860f-7ef07be22394 https://doi.org/10.1109/TIE.2025.3579075 |
| Rights: | IEEE Copyright |
| Přístupové číslo: | edsair.doi.dedup.....2e71ad91bc817640a3de16832e550503 |
| Databáze: | OpenAIRE |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstrakt: | Multirobot formation planning has wide applications across various domains. This article proposes an innovative neural-controller-based approach to address the formation planning problem. The proposed noise-tolerant fixed-time zeroing neural network (NT-FTZNN) controller achieves convergence under constant noise, dynamic bounded noise, and even dynamic unbounded noise, demonstrating strong adaptability in complicated scenarios. To the best of our knowledge, this article is the first application of a neural controller with both noise robustness and fixed-time convergence properties to multirobot formation planning tasks. In the presence of various kinds of noises, the proposed method achieves a formation error on the order of 10−7, which significantly outperforms other advanced formation control methods that typically reach only the 10−2 level. Moreover, rigorous theoretical analysis proves that the proposed controller guarantees global stability and fixed-time convergence under various noise conditions. Extensive numerical simulations and physical experiments further validate the superiority of the proposed approach over existing methods, confirming its practical effectiveness in real-world multirobot formation tasks. |
|---|---|
| ISSN: | 15579948 02780046 |
| DOI: | 10.1109/tie.2025.3579075 |