Latent Linear Quadratic Regulator for Robotic Control Tasks
This addresses computational bottlenecks in robotic control for applications requiring real-time performance, representing an incremental improvement by adapting existing methods to reduce complexity.
The paper tackles the high computational cost of model predictive control (MPC) in robotic tasks by introducing a latent linear quadratic regulator (LaLQR) that maps states to a latent space with linear dynamics and quadratic costs, enabling efficient LQR application. Experiments demonstrate LaLQR achieves superior efficiency and generalization over baselines, though no specific numerical results are provided.
Model predictive control (MPC) has played a more crucial role in various robotic control tasks, but its high computational requirements are concerning, especially for nonlinear dynamical models. This paper presents a $\textbf{la}$tent $\textbf{l}$inear $\textbf{q}$uadratic $\textbf{r}$egulator (LaLQR) that maps the state space into a latent space, on which the dynamical model is linear and the cost function is quadratic, allowing the efficient application of LQR. We jointly learn this alternative system by imitating the original MPC. Experiments show LaLQR's superior efficiency and generalization compared to other baselines.