LQGNet: Hybrid Model-Based and Data-Driven Linear Quadratic Stochastic Control
This work addresses a key limitation in stochastic control for applications where system dynamics are not fully known, offering an incremental improvement over existing LQG methods.
The paper tackles the problem of stochastic control with partially known linear Gaussian state-space models, presenting LQGNet, a hybrid model-based and data-driven controller that outperforms classic methods by overcoming model mismatches, as shown empirically.
Stochastic control deals with finding an optimal control signal for a dynamical system in a setting with uncertainty, playing a key role in numerous applications. The linear quadratic Gaussian (LQG) is a widely-used setting, where the system dynamics is represented as a linear Gaussian statespace (SS) model, and the objective function is quadratic. For this setting, the optimal controller is obtained in closed form by the separation principle. However, in practice, the underlying system dynamics often cannot be faithfully captured by a fully known linear Gaussian SS model, limiting its performance. Here, we present LQGNet, a stochastic controller that leverages data to operate under partially known dynamics. LQGNet augments the state tracking module of separation-based control with a dedicated trainable algorithm. The resulting system preserves the operation of classic LQG control while learning to cope with partially known SS models without having to fully identify the dynamics. We empirically show that LQGNet outperforms classic stochastic control by overcoming mismatched SS models.