Abstract:We introduce a compositional data-driven methodology for designing fully-decentralized safety controllers applicable to large-scale interconnected networks, encompassing subsystems with unknown mathematical models. Our compositional scheme leverages the interconnection topology and breaks down the network analysis into the examination of distinct subsystems. This is accompanied by utilizing a concept of control storage certificates (CSCs) to capture joint dissipativity-type properties among subsystems. These CSCs are instrumental in a compositional derivation of a control barrier certificate (CBC) specialized for the interconnected network, thereby ensuring its safety. In our data-driven scheme, we gather solely one input-output trajectory from each unknown subsystem within a specified time frame. By fulfilling a specific rank condition, this process facilitates the construction of a CSC for each subsystem. Following this, by adhering to compositional dissipativity reasoning, we compose CSCs derived from data and build a CBC for the unknown network, ensuring its safety over an infinite time horizon, while providing correctness guarantees. We demonstrate that our compositional data-driven approach significantly enhances the design of a CBC and its safety controller across the interconnected network. This advancement is achieved by reducing the computational complexity from a polynomial growth in relation to network dimension, when using sum-of-squares (SOS) optimization, to a linear scale based on the number of subsystems. We additionally demonstrate that the dissipativity-type compositionality condition can benefit from the structure of interconnection topology and potentially be fulfilled regardless of the number of subsystems. We apply our data-driven findings to a variety of benchmarks, involving physical networks with unknown models and diverse interconnection topologies.

Data-Driven Control of Large-Scale Networks with Formal Guarantees: A Small-Gain Free Approach

From Data to Control: A Formal Compositional Framework for Large-Scale Interconnected Networks

Symbolic Abstractions with Guarantees: A Data-Driven Divide-and-Conquer Strategy

Abstract interpretation of distributed network control planes

Formal Control of New England 39-Bus Test System: An Assume-Guarantee Approach

Data-driven Construction of Finite Abstractions for Interconnected Systems: A Compositional Approach

A Data-Driven Safety Preserving Control Architecture for Constrained Cyber-Physical Systems

Cooperative Control with Distributed Gain Adaptation and Connectivity Estimation for Directed Networks

Controlling Directed Networks with Evolving Topologies

A data‐driven safety preserving control architecture for constrained cyber‐physical systems

From Small-Gain Theory to Compositional Construction of Barrier Certificates for Large-Scale Stochastic Systems

Data-driven abstraction-based control synthesis

Distributed Synthesis of Local Controllers for Networked Systems with Arbitrary Interconnection Topologies

Control of Networked Multiagent Systems with Uncertain Graph Topologies

Prevalence and scalable control of localized networks

A purely data-driven framework for prediction, optimization, and control of networked processes

Compositional Design of Safety Controllers for Large-scale Stochastic Hybrid Systems

Abstraction-based Control of Unknown Continuous-Space Models with Just Two Trajectories

Data-Driven Resilient Control for Linear Discrete-Time Multi-Agent Networks under Unconfined Cyber-Attacks

Compositional Neural Certificates for Networked Dynamical Systems

Layered Affine Formation Control of Networked Uncertain Systems: A Fully Distributed Approach Over Directed Graphs