Local Trajectory Optimization

Local trajectory optimization focuses on efficiently finding optimal robot movement paths in complex or dynamic environments, aiming to minimize costs like energy consumption or travel time while satisfying constraints. Current research emphasizes methods combining local optimization algorithms like Model Predictive Path Integral (MPPI) and iterative Linear Quadratic Regulators (iLQR) with global guidance from techniques such as graph search, Gaussian Processes, or learned models (e.g., using temporal difference learning). These advancements enable improved performance in challenging scenarios, including multi-robot systems and human-robot interaction, with applications ranging from autonomous vehicles to robotic manipulation.