arXiv cs.MA· Panagiotis Kasnesis, Lazaros Toumanidis, Christos Chatzigeorgiou, Amalia Contiero Syropoulou, Miguel De Prado·· 3 天前AI 评分24
每区一智能体：部分可观测下多区 HVAC 的边缘去中心化主动推理控制
One Agent per Zone: Decentralised Active Inference under Partial Observability for Multi-Zone HVAC Control on Edge Hardware
AI 导读
研究提出一种去中心化控制器，为每个区域分配一个独立的主动推理智能体，通过最小化期望自由能规划并在线学习区域动态，智能体之间不交换消息。在 Sinergym 办公楼 EnergyPlus 仿真中，15 个智能体在五个未见天气年份上实现最低建筑峰值 88.3 kW（比基线低 3% 至 17%），同时系数 0.929，优于 0.936 至 0.994 的基线。
正文 · 原文
Abstract:Multi-zone HVAC control is a shared-resource problem: zones on a floor share one fan and cooling coil, so each zone's demand affects its neighbours and the building's peak power. One controller per zone scales well and isolates faults, but each controller then observes only its own zone, not the shared air loop. In this work, we present a decentralised controller that assigns each zone an independent active-inference agent, which plans by minimising expected free energy and learns its zone dynamics online. The agents exchange no messages, so none depends on peers or a coordinator; each accounts for the unobserved coupling through a latent disturbance and a congestion signal derived from the metered power of the other floors. Fifteen agents are evaluated in EnergyPlus on a Sinergym office building against three rule-based controllers and three MADDPG fleets, one given the same floor meters. On five noisy weather years unseen during training, the fleet attains the lowest mean building peak (88.3 kW, 3 to 17 % below the baselines) and staggers the floor peaks better than any baseline, with a coincidence factor (building peak over the sum of the floor peaks) of 0.929 versus 0.936 to 0.994. Compared with a reactive rule at similar comfort (0.912 versus 0.916), it lowers the sum of monthly peaks, on which demand charges are billed, by 33 kW (4 %) in every year, at the cost of about 10 % more energy. It costs 9% less per year than the proactive rule (at lower comfort) and 1% less than the MADDPG fleet given the same floor meters, in every test year. The whole fleet runs on a single low-cost Raspberry Pi 5, closing the observation-to-action loop over MQTT in 442 ms at p95, far below the 15-minute control step.
| Comments: | 9 pages, 3 figures, accepted in European Conference on EDGE AI Technologies and Applications, EEAI 2026, Malta |
| Subjects: | Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.19790 [cs.MA] |
| (or arXiv:2610.19790v1 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2610.19790 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Panagiotis Kasnesis [view email]
[v1]
Wed, 14 Oct 2026 08:36:32 UTC (69 KB)
来源：arXiv cs.MA · arxiv.org