A Separation-Based Design to Data-Driven Control for Large-Scale Partially Observed Systems

TL;DR

This paper introduces a separation-based approach for data-driven control of large-scale partially observed systems governed by PDEs, combining trajectory optimization with LQG control design based on experimental data.

Contribution

It proposes a novel separation-based method that integrates trajectory optimization and data-driven LQG control for complex PDE-governed systems.

Findings

Effective control demonstrated on a nonlinear heat example

Combines black box simulation with experimental data for control design

Addresses large-scale PDE systems with partial observations

Abstract

This paper studies the partially observed stochastic optimal control problem for systems with state dynamics governed by Partial Differential Equations (PDEs) that leads to an extremely large problem. First, an open-loop deterministic trajectory optimization problem is solved using a black box simulation model of the dynamical system. Next, a Linear Quadratic Gaussian (LQG) controller is designed for the nominal trajectory-dependent linearized system, which is identified using input-output experimental data consisting of the impulse responses of the optimized nominal system. A computational nonlinear heat example is used to illustrate the performance of the approach.