Consensus Tracking Data-Driven Control of Multi-Agent Systems based on Matrix S-Lemma Under Noisy Data
Shuli Tan, Xiufeng Zhang, Chunxi Yang · 2024
In this paper, the problem of consensus tracking for multi-agent systems in the presence of noise inter-ference is investigated. Unlike the traditional model-based approach, this paper assumes that the dynamics of all agents are unknown and only a portion of the input-state data is available. The offline data contaminated with noise interference is utilized. By introducing the matrix S-lemma, a necessary and sufficient conditions for directly designing consensus tracking controllers from bounded noisy input-state data is given. This method can be applied to large datasets and effectively addresses computational challenges associated with extensive data. Simulation result illustrates the effectiveness of the theoretical analysis.