Model Predictive Trees: Sample-Efficient Receding Horizon Planning with Reusable Tree Search

TL;DR

Model Predictive Trees (MPT) is a novel receding horizon planning algorithm that reuses entire optimal subtrees to improve efficiency and performance in dynamic environments, outperforming existing methods.

Contribution

The paper introduces MPT, a new tree search method that reuses full subtrees for better planning efficiency and accuracy in dynamic scenarios.

Findings

Outperforms state-of-the-art sampling-based methods

Effectively guides search away from low-quality areas

Demonstrated on autonomous vehicle manipulation task

Abstract

We present Model Predictive Trees (MPT), a receding horizon tree search algorithm that improves its performance by reusing information efficiently. Whereas existing solvers reuse only the highest-quality trajectory from the previous iteration as a "hotstart", our method reuses the entire optimal subtree, enabling the search to be simultaneously guided away from the low-quality areas and towards the high-quality areas. We characterize the restrictions on tree reuse by analyzing the induced tracking error under time-varying dynamics, revealing a tradeoff between the search depth and the timescale of the changing dynamics. In numerical studies, our algorithm outperforms state-of-the-art sampling-based cross-entropy methods with hotstarting. We demonstrate our planner on an autonomous vehicle testbed performing a nonprehensile manipulation task: pushing a target object through an obstacle field. Code associated with this work will be made available at https://github.com/jplathrop/mpt.