Partner Institution: European Commission

Project Information: Co-Principal Investigator (Co-PI), €1.85 million, 2010.01-2013.11

Project Description: Existing theoretical and practical tools cannot scalably and efficiently handle control problems for complex, uncertain, and time-varying large systems, which not only leads to time-consuming, labor-intensive, and costly deployment of Large-Scale Control Systems (LSCSs) but also hinders the widespread application of LSCSs in fields where they could significantly improve system efficiency and Quality of Service (QoS), reduce energy consumption and emissions, and improve quality of daily life. Building on its partners’ latest advances in LSCS convex optimization design and robust, efficient LSCS self-tuning, the AGILE project aims to develop and evaluate an integrated LSCS design method applicable to large systems of any scale, heterogeneity, and complexity, capable of: - providing proactive, arbitrarily close-to-optimal LSCS performance; - having inherent self-tuning capabilities to quickly and effectively optimize LSCS performance when short-term, medium-term, and long-term changes affect large systems; - providing efficient, fast, and safe fault recovery and LSCS reconfiguration; and, - achieving all of the above while maintaining scalability and modularity. To facilitate implementation and deployment of the AGILE system in existing open architecture SCADA/DCS infrastructure, a set of open-source interface tools will be developed. The integrated LSCS design system and its interfaces developed in the AGILE project will be extensively tested and evaluated in two real large-scale test cases (an urban traffic network with 20 intersections and a large energy-controlled building). These two cases feature rich design and performance characteristics, extremely complex nonlinear dynamics, high stochastic effects, uncertainty and modeling errors, and requirements for reconfiguration and modular design.