Data-Driven System Level Synthesis
This work provides a method for control system design without explicit model identification, which is significant for engineers working with complex systems where models are difficult to obtain.
This paper develops data-driven versions of System Level Synthesis (SLS) for linear-time-invariant systems over a finite horizon, enabling optimization over system responses using only past system trajectories without explicit model identification. It demonstrates an exact equivalence between traditional and data-driven SLS in a noise-free setting and shows how robust SLS and trajectory averaging can mitigate the effects of process noise.
We establish data-driven versions of the System Level Synthesis (SLS) parameterization of achievable closed-loop system responses for a linear-time-invariant system over a finite-horizon. Inspired by recent work in data-driven control that leverages tools from behavioral theory, we show that optimization problems over system-responses can be posed using only libraries of past system trajectories, without explicitly identifying a system model. We first consider the idealized setting of noise free trajectories, and show an exact equivalence between traditional and data-driven SLS. We then show that in the case of a system driven by process noise, tools from robust SLS can be used to characterize the effects of noise on closed-loop performance, and further draw on tools from matrix concentration to show that a simple trajectory averaging technique can be used to mitigate these effects. We end with numerical experiments showing the soundness of our methods.