Work

Balanced Passive System Reduction for Networked Robotic Manipulators in Multi-Agent Systems

Year: 2025
Type: article
Abstract: This paper introduces a novel application of the Balanced Passive System Reduction (BPSR) algorithm to a networked multi-agent system of six 4 DOF robotic manipulators interconnected in a leader–follower consensus topology. Unlike classical balanced truncation or moment matching methods, BPSR preserves both passivity and Lyapunov based stability during reduction, ensuring that the reduced models remain physically meaningful. We reduce the original 48th order LTI model to third order and fourth order approximations and quantify their accuracy using H∞ and H2 norms the order 4 model achieves error norms of 0.00294 and 0.00276, respectively, compared to 0.01998 and 0.01160 for the order 3 model. We also present explicit stability and passivity proofs to bolster reproducibility. Frequency domain (Nichols, Nyquist, Bode, and passive) and time domain (step) simulations confirm that the fourth order model closely replicates the original dynamics. By reducing computational complexity while maintaining performance, BPSR enables robust real time control in multi agent robotic applications, addressing both scalability and controller design requirements.
Language: English
Open Access status: diamond
OpenAlex ID: W7160912055
FWCI: 0.227
Cites: 79
Cited by: 1
Related to: 10