Improving Kinodynamic Planners for Vehicular Navigation with Learned Goal-Reaching Controllers

10/08/2021
by   Aravind Sivaramakrishnan, et al.
0

This paper aims to improve the path quality and computational efficiency of sampling-based kinodynamic planners for vehicular navigation. It proposes a learning framework for identifying promising controls during the expansion process of sampling-based planners. Given a dynamics model, a reinforcement learning process is trained offline to return a low-cost control that reaches a local goal state (i.e., a waypoint) in the absence of obstacles. By focusing on the system's dynamics and not knowing the environment, this process is data-efficient and takes place once for a robotic system. In this way, it can be reused in different environments. The planner generates online local goal states for the learned controller in an informed manner to bias towards the goal and consecutively in an exploratory, random manner. For the informed expansion, local goal states are generated either via (a) medial axis information in environments with obstacles, or (b) wavefront information for setups with traversability costs. The learning process and the resulting planning framework are evaluated for a first and second-order differential drive system, as well as a physically simulated Segway robot. The results show that the proposed integration of learning and planning can produce higher quality paths than sampling-based kinodynamic planning with random controls in fewer iterations and computation time.

READ FULL TEXT

page 1

page 3

research
01/06/2022

Data-Efficient Learning of High-Quality Controls for Kinodynamic Planning used in Vehicular Navigation

This paper aims to improve the path quality and computational efficiency...
research
07/18/2019

Towards Learning Efficient Maneuver Sets for Kinodynamic Motion Planning

Planning for systems with dynamics is challenging as often there is no l...
research
04/19/2022

Learning Forward Dynamics Model and Informed Trajectory Sampler for Safe Quadruped Navigation

For autonomous quadruped robot navigation in various complex environment...
research
04/11/2021

Guided Incremental Local Densification for Accelerated Sampling-based Motion Planning

Sampling-based motion planners rely on incremental densification to disc...
research
07/26/2023

Improving Reliable Navigation under Uncertainty via Predictions Informed by Non-Local Information

We improve reliable, long-horizon, goal-directed navigation in partially...
research
05/22/2018

Global Navigation Using Predictable and Slow Feature Analysis in Multiroom Environments, Path Planning and Other Control Tasks

Extended Predictable Feature Analysis (PFAx) [Richthofer and Wiskott, 20...
research
11/06/2022

Depth-based Sampling and Steering Constraints for Memoryless Local Planners

By utilizing only depth information, the paper introduces a novel but ef...

Please sign up or login with your details

Forgot password? Click here to reset