MINER-RRT*: A Hierarchical and Fast Trajectory Planning Framework in 3D Cluttered Environments

TL;DR

MINER-RRT* is a hierarchical 3D trajectory planning framework for quadrotors that combines neural network-boosted sampling and differential flatness-based polynomial trajectories to achieve faster, more efficient path generation in cluttered environments.

Contribution

The paper introduces MINER-RRT*, a novel hierarchical framework that significantly improves planning speed and efficiency using neural network heuristics and differential flatness.

Findings

Faster trajectory generation compared to SOTA methods.

High-quality trajectories in complex 3D cluttered environments.

Validated through extensive simulations and real-world experiments.

Abstract

Trajectory planning for quadrotors in cluttered environments has been challenging in recent years. While many trajectory planning frameworks have been successful, there still exists potential for improvements, particularly in enhancing the speed of generating efficient trajectories. In this paper, we present a novel hierarchical trajectory planning framework to reduce computational time and memory usage called MINER-RRT*, which consists of two main components. First, we propose a sampling-based path planning method boosted by neural networks, where the predicted heuristic region accelerates the convergence of rapidly-exploring random trees. Second, we utilize the optimal conditions derived from the quadrotor's differential flatness properties to construct polynomial trajectories that minimize control effort in multiple stages. Extensive simulation and real-world experimental results demonstrate that, compared to several state-of-the-art (SOTA) approaches, our method can generate high-quality trajectories with better performance in 3D cluttered environments.