2025/11/11 by Li, Shihao, Li, Jiachen, Xu, Jiamin +1 · 1 citation
#FOS: Electrical engineering #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · doi:10.48550/arxiv.2511.07894
We present S2C (Specification-to-Certified-Controller), a multi-agent framework that maps natural-language requirements to certified H_∞ state-feedback controllers via LMI synthesis. S2C coordinates five roles -- SpecInt (spec extraction), Solv (bounded-real lemma (BRL) LMI), Tester (Monte Carlo and frequency-domain checks), Adapt (spec refinement), and CodeGen (deployable code). The loop is stabilized by a severity- and iteration-aware γ-floor guardrail and a decay-rate region constraint enforcing \Reλ(A+BK)<-α with α=3.9/Ts derived from settling-time targets. For state feedback, verification reports disturbance rejection ‖C (sI-(A+BK))-1E‖_∞ alongside time-domain statistics; discrete benchmarks are converted to continuous time via a Tustin (bilinear) transform when needed. On 14 COMPleib problems, S2C attains 100% synthesis success and 100% convergence within six iterations, with strong decay-rate satisfaction and near-target certified H_∞ levels; it improves robustness metrics relative to single-shot BRL and BRL+α baselines. An ablation over LLM backbones (GPT-5, GPT-5 mini, DeepSeek-V3, Qwen-2.5-72B, Llama-4 Maverick) shows the pipeline is robust across models, while stronger models yield the highest effectiveness. These results indicate that LLM agents can integrate certificate-bearing control synthesis from high-level intent, enabling rapid end-to-end prototyping without sacrificing formal guarantees.