RRT+ : Fast Planning for High-Dimensional Configuration Spaces (original) (raw)

RRT-connect: An efficient approach to single-query path planning

2000

A simple and efficient randomized algorithm is presented for solving single-query path planning problems in high-dimensional configuration spaces. The method works by incrementally building two Rapidly-exploring Random Trees (RRTs) rooted at the start and the goal configurations. The trees each explore space around them and also advance towards each other through the use of a simple greedy heuristic. Although originally designed to plan motions for a human arm (modeled as a 7-DOF kinematic chain) for the automatic graphic animation of collision-free grasping and manipulation tasks, the algorithm has been successfully applied to a variety of path planning problems. Computed examples include generating collision-free motions for rigid objects in 2D and 3D, and collision-free manipulation motions for a 6-DOF PUMA arm in a 3D workspace. Some basic theoretical analysis is also presented.

An adaptive roadmap guided Multi-RRTs strategy for single query path planning

2010 IEEE International Conference on Robotics and Automation, 2010

During the past decade, Rapidly-exploring Random Tree (RRT) and its variants are shown to be powerful sampling based single query path planning approaches for robots in high-dimensional configuration space. However, the performance of such tree-based planners that rely on uniform sampling strategy degrades significantly when narrow passages are contained in the configuration space. Given the assumption that computation resources should be allocated in proportion to the geometric complexity of local region, we present a novel single-query Multi-RRTs path planning framework that employs an improved Bridge Test algorithm to identify global important roadmaps in narrow passages. Multiple trees can be grown from these sampled roadmaps to explore sub-regions which are difficult to reach. The probability of selecting one particular tree for expansion and connection, which can be dynamically updated by on-line learning algorithm based on the historic results of exploration, guides the tree through narrow passage rapidly. Experimental results show that the proposed approach gives substantial improvement in planning efficiency over a wide range of single-query path planning problems. I. INTRODUCTION OBOT path planning has been one of the fundamental problems over the last couple of decades in such areas as robotics, artificial intelligence, as well as computer graphics. The original description of the problem is to plan a collision-free path for a robot made of an arbitrary number of polyhedral bodies among an arbitrary number of polyhedral obstacles between two collision-free queried positions of the robot, which has been shown to be PSPACE-complete by complex geometric analysis [1]. The well known complete motion planning algorithms, such as cell decomposition and visibility roadmaps, require explicit representation of robot configuration space. They are usually computationally intractable and hard to implement for practical applications [2]. "The curse of dimensionality" has lead to the development of randomized sampling-based motion planners, which can solve many previously considered hard problems successfully and quickly. PRM [3] and RRT [4] are two typical randomized Manuscript received

RRT-HX: RRT With Heuristic Extend Operations for Motion Planning in Robotic Systems

Volume 5A: 40th Mechanisms and Robotics Conference, 2016

This paper presents a sampling-based method for path planning in robotic systems without known cost-to-go information. It uses trajectories generated from random search to heuristically learn the cost-to-go of regions within the configuration space. Gradually, the search is increasingly directed towards lower cost regions of the configuration space, thereby producing paths that converge towards the optimal path. The proposed framework builds on Rapidly-exploring Random Trees for random sampling-based search and Reinforcement Learning is used as the learning method. A series of experiments were performed to evaluate and demonstrate the performance of the proposed method.

Motion Planning by Sampling in Subspaces of Progressively Increasing Dimension

Journal of Intelligent & Robotic Systems

This paper introduces an enhancement to traditional sampling-based planners, resulting in efficiency increases for high-dimensional holonomic systems such as hyperredundant manipulators, snake-like robots, and humanoids. Despite the performance advantages of modern sampling-based motion planners, solving high dimensional planning problems in near real-time remains a considerable challenge. The proposed enhancement to popular sampling-based planning algorithms is aimed at circumventing the exponential dependence on dimensionality, by progressively exploring lower dimensional volumes of the configuration space. Extensive experiments comparing the enhanced and traditional version of RRT, RRT-Connect, and Bidirectional T-RRT on both a planar hyper-redundant manipulator and the Baxter humanoid robot show significant acceleration, up to two orders of magnitude, on computing a solution. We also explore important implementation issues in the sampling process and discuss the limitations of this method.

Informed RRT*-Connect: An Asymptotically Optimal Single-Query Path Planning Method

IEEE Access

Rapidly-exploring Random Trees (RRTs) are successful in single-query motion planning problems. The standard version of RRT grows a tree from a start location and stops once it reached the goal configuration. RRT-Connect is the bidirectional version of RRT, which grows two trees simultaneously. These two trees try to establish a connection to stop searching. RRT-Connect finds solutions faster than RRT. Following that, an asymptotically optimal version of RRT-Connect called RRT*-Connect has been introduced. It not only rewires both trees while they are growing, but also it keeps searching the state space for better solutions than the current one. However, it is inefficient and inconsistent to search all over the state space in order to find better solutions than the current one concerning its single-query nature. The better way is to look through states that can provide a better solution. In this paper, we propose Informed RRT*-Connect, which is the informed version of RRT*-Connect that uses direct sampling after the first solution found. Unlike RRT*-Connect, the proposed method checks only the states that can potentially provide better solutions than the current solution. The proposed method benefited from the properties of RRT*-Connect and informed sampling, which offers low-cost solutions with fewer iterations in comparison to RRT*-Connect. Different simulations in OMPL have been carried out to show the significance of Informed RRT*-Connect in comparison with RRT*, Informed RRT*, and RRT*-Connect.

Preprocessing of Configuration Space for Improved Sampling Based Path Planning

Sampling based planners have been successful in path planning of robots with many degrees of freedom, but still remains ineffective when the configuration space has a narrow passage. This paper presents two new techniques of preprocessing the configuration space. The first technique called a Random Walk to Surface (RWS), uses a random walk strategy to generate samples in narrow regions quickly, thus improving effciency of Probabilistic Roadmap (PRM) based planners. The algorithm substantially reduces instances of collision checking and thereby decreases computational time. The method is powerful even for cases where the structure of the narrow passage is not known a priori, thus giving significant improvement over other known methods. The second method, by preprocessing the configuration space, improves the effiency of Rapidly Exploring Random Tree (RRT) like planners by identifying key regions of the configuration space to search for a solution path. The Experiments show a significant improvement in effiency for both PRM and RRT like planners.

A scalable distributed RRT for motion planning

2013 IEEE International Conference on Robotics and Automation, 2013

Rapidly-exploring Random Tree (RRT), like other sampling-based motion planning methods, has been very successful in solving motion planning problems. Even so, samplingbased planners cannot solve all problems of interest efficiently, so attention is increasingly turning to parallelizing them. However, one challenge in parallelizing RRT is the global computation and communication overhead of nearest neighbor search, a key operation in RRTs. This is a critical issue as it limits the scalability of previous algorithms. We present two parallel algorithms to address this problem. The first algorithm extends existing work by introducing a parameter that adjusts how much local computation is done before a global update. The second algorithm radially subdivides the configuration space into regions, constructs a portion of the tree in each region in parallel, and connects the subtrees, removing cycles if they exist. By subdividing the space, we increase computation locality enabling a scalable result. We show that our approaches are scalable. We present results demonstrating almost linear scaling to hundreds of processors on a Linux cluster and a Cray XE6 machine.

A REAL-TIME MOTION PLANNING ALGORITHM FOR A HYPER-REDUNDANT SET OF MECHANISMS

We introduce a novel probabilistic algorithm (CPRM) for real-time motion planning in the configuration space C. Our algorithm differs from a Probabilistic Road Map algorithm (PRM) in the motion between a pair of anchoring points (local planner) which takes place on the boundary of the obstacle subspace O. We define a varying potential field f on ∂O as a Morse function and follow ∇f . We then exemplify our algorithm on a redundant worm climbing robot with n degrees of freedom and compare our algorithm running results with those of PRM.

A scalable method for parallelizing sampling-based motion planning algorithms

2012

Abstract—This paper describes a scalable method for paral-lelizing sampling-based motion planning algorithms. It subdi-vides configuration space (C-space) into (possibly overlapping) regions and independently, in parallel, uses standard (sequen-tial) sampling-based planners to construct roadmaps in each region. Next, in parallel, regional roadmaps in adjacent regions are connected to form a global roadmap. By subdividing the space and restricting the locality of connection attempts, we reduce the work and inter-processor communication associated with nearest neighbor calculation, a critical bottleneck for scalability in existing parallel motion planning methods. We show that our method is general enough to handle a variety of planning schemes, including the widely used Probabilistic Roadmap (PRM) and Rapidly-exploring Random Trees (RRT) algorithms. We compare our approach to two other existing parallel algorithms and demonstrate that our approach achieves better and more scalable pe...