Learning Space Partitions for Path Planning

06/19/2021
by   Kevin Yang, et al.
5

Path planning, the problem of efficiently discovering high-reward trajectories, often requires optimizing a high-dimensional and multimodal reward function. Popular approaches like CEM and CMA-ES greedily focus on promising regions of the search space and may get trapped in local maxima. DOO and VOOT balance exploration and exploitation, but use space partitioning strategies independent of the reward function to be optimized. Recently, LaMCTS empirically learns to partition the search space in a reward-sensitive manner for black-box optimization. In this paper, we develop a novel formal regret analysis for when and why such an adaptive region partitioning scheme works. We also propose a new path planning method PlaLaM which improves the function value estimation within each sub-region, and uses a latent representation of the search space. Empirically, PlaLaM outperforms existing path planning methods in 2D navigation tasks, especially in the presence of difficult-to-escape local optima, and shows benefits when plugged into model-based RL with planning components such as PETS. These gains transfer to highly multimodal real-world tasks, where we outperform strong baselines in compiler phase ordering by up to 245 properties on a 0-1 scale.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/01/2020

Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Search

High dimensional black-box optimization has broad applications but remai...
research
05/03/2018

Two Techniques That Enhance the Performance of Multi-robot Prioritized Path Planning

We introduce and empirically evaluate two techniques aimed at enhancing ...
research
02/28/2019

Learning to Plan via Neural Exploration-Exploitation Trees

Sampling-based algorithms such as RRT and its variants are powerful tool...
research
02/25/2021

LES: Locally Exploitative Sampling for Robot Path Planning

Sampling-based algorithms solve the path planning problem by generating ...
research
09/13/2019

Deep Learned Path Planning via Randomized Reward-Linked-Goals and Potential Space Applications

Space exploration missions have seen use of increasingly sophisticated r...
research
10/31/2011

Exploiting Subgraph Structure in Multi-Robot Path Planning

Multi-robot path planning is difficult due to the combinatorial explosio...
research
12/15/2021

Enhance Connectivity of Promising Regions for Sampling-based Path Planning

Sampling-based path planning algorithms usually implement uniform sampli...

Please sign up or login with your details

Forgot password? Click here to reset