标题：Determining dissipation inequalities from input-output samples * * The authors thank the German Research Foundation (DFG) for financial support of the project within the Cluster of Excellence in Simulation Technology (EXC 310/2) at the University of Stuttgart.
摘要：AbstractKnowledge of a dissipation inequality can disclose valuable system properties relevant for controller design. Therefore, we consider the problem of determining a dissipation inequality for an input-output system. In practice, yet, the core issue is that one would only know finitely many input-output samples while the input-output relationship would be undisclosed analytically. For such scenarios, we provide an approach for computing dissipation inequalities based upon knowledge of large amounts of input-output samples, as they arise in applications nowadays gathered under the term “big data”. Our approach will, under certain conditions, provide dissipation inequalities which remain satisfied for all input-output pairs that the system can produce, though only having been derived from finitely many of them.
