Multi-Strategy Improved Rapid Random Expansion Tree (RRT) Algorithm for Robotic Arm Path Planning

Yuan Sun, Shoujun ZhangInternational Journal of Advanced Computer Science and Applications 2025 · DOI

TL;DR

This paper introduces a "Multi-Strategy Improved RRT" algorithm that significantly enhances robotic arm path planning by combining bidirectional tree growth, intelligent sampling, and adaptive step sizes. By integrating these techniques with cubic B-spline smoothing, the researchers achieved a staggering 96% reduction in planning time and an 18% shorter path length compared to traditional RRT in complex environments.

What Problem Does This Paper Address?

Path planning is the "bread and butter" of robotics, yet the classic Rapidly-exploring Random Tree (RRT) algorithm—a staple in the field—has notorious drawbacks. Because it relies on purely random sampling, it often behaves like a "drunkard’s walk," resulting in:

  1. Low Efficiency: It spends too much time exploring irrelevant areas.
  2. Redundancy: It generates a massive number of unnecessary nodes.
  3. Poor Path Quality: The resulting paths are often jagged and "long-winded," making them difficult for physical robotic arms to execute smoothly without causing mechanical wear or instability.

The authors aim to solve these issues, particularly for robotic arms operating in narrow or complex 3D spaces where precision and speed are non-negotiable.

Methods

The authors don't just tweak one knob; they overhaul the RRT framework using four synergistic strategies:

  • Bidirectional Tree Extension: Instead of growing one tree from the start to the goal, the algorithm grows two trees simultaneously—one from the start and one from the target. They "meet in the middle," which drastically cuts down search time.
  • Target-Biased & Local Gaussian Sampling: To stop the algorithm from wandering aimlessly, they introduce a probability-based bias toward the goal. They also use Gaussian distribution sampling to concentrate search points more densely as the tree gets closer to the target, ensuring a more focused "final approach."
  • Adaptive Step Size: Traditional RRT uses a fixed step size. This algorithm uses a "zoom" logic: it takes large steps in open areas to cover ground quickly and smaller, more precise steps when it's near the goal or navigating tight obstacles.
  • Path Smoothing (Cubic B-splines & Minimum Curvature): Raw RRT paths are jagged. The authors apply cubic B-spline interpolation and curvature optimization to turn these sharp corners into smooth, continuous curves that a robotic arm can actually follow.

Datasets

The study relies on custom-designed simulation environments created in MATLAB R2021b. The authors tested the algorithm across four distinct scenarios:

  1. Normal Map: Standard obstacle density.
  2. Narrow Map: Featuring extremely tight passages to test "bottleneck" navigation.
  3. Complex Map: High-density obstacles to test search robustness.
  4. 3D Map: A 800×800×800 voxel space to simulate real-world robotic arm reach.

Each scenario was run 100 times (20 times for the 3D map) to ensure statistical significance.

Key Insights and Ideas

  • The Power of "Meeting in the Middle": The bidirectional approach (RRT-Connect style) is the heavy lifter for speed, but it's the combination with adaptive steps that prevents the two trees from "missing" each other in complex corridors.
  • Balancing Exploration and Exploitation: By adjusting the Prand value (target bias), the algorithm can dynamically balance between searching the whole room and sprinting for the finish line.
  • Smoothness is Feasibility: In robotics, a path isn't just a line; it's a set of motor commands. The transition from a "node-based" path to a "curvature-optimized" path is what makes this research practical for industrial applications.

Key Results

The improvements over the traditional RRT algorithm are dramatic, especially in the Complex Map mode:

  • Planning Time: Reduced by 96.29% (from ~10 seconds down to 0.37 seconds).
  • Path Length: Shortened by 18.15%.
  • Memory/Compute Efficiency: The number of nodes required dropped by 92.13%, and iterations fell by 91.60%.

In 3D environments, the results were even more pronounced. The improved RRT outperformed RRT* and RRT-Connect across every metric, proving that the multi-strategy approach scales effectively to higher dimensions.

Limitations and Future Work

While the results are impressive, the authors acknowledge a few areas for growth:

  • Static Environments: The current simulations focus on static obstacles. In a real-world factory or hospital, obstacles (like humans or other robots) move.
  • Dimensionality: While they tested in 3D, many robotic arms have 6 or 7 Degrees of Freedom (DoF). Extending this logic to higher-dimensional configuration spaces is the next logical step.
  • Future Direction: The authors plan to extend the algorithm to "dynamic multi-dimensional obstacle environments," which would allow robotic arms to navigate safely in changing, real-time settings.

Summary updated

Abstract

The purpose of this paper is to propose an improved RRT algorithm that incorporates multiple improvement strategies to solve the problems of low efficiency, long and unsmooth paths in the traditional rapid random expansion tree (RRT) algorithm for path planning of robotic arms. The algorithm first uses a bidirectional tree extension strategy to generate trees from both the starting point and the target position simultaneously, improving search efficiency and reducing redundant paths. Secondly, the algorithm introduces target bias sampling in combination with local Gaussian sampling, which renders the sampling points more focused on the target area, and dynamically adjusts the distribution to improve sampling efficiency and path connection speed. Concurrently, the algorithm is equipped with an adaptive step size strategy, which dynamically adjusts the expansion step size according to the target distance, thereby achieving a balance between rapid expansion over long distances and precise search at close range. Finally, a collision-free operation is ensured by a path verification mechanism, and the path is smoothed using cubic B-splines and minimum curvature optimisation techniques, significantly improving the smoothness of the path and the feasibility of the robot arm movement. As demonstrated by simulation experiments, the improved RRT algorithm exhibits a reduction in the average path length by 18.15%, planning time by 96.29%, the number of nodes by 92.13%, and the number of iterations by 91.60%, in comparison with the conventional RRT algorithm, when operating in complex map mode. These findings substantiate the efficacy and practicality of the improved RRT algorithm in the domain of robotic arm path planning.
BibTeX
@article{sun2025multi,
  author = {Yuan Sun and Shoujun Zhang},
  title = {{Multi-Strategy Improved Rapid Random Expansion Tree (RRT) Algorithm for Robotic Arm Path Planning}},
  journal = {International Journal of Advanced Computer Science and Applications},
  year = {2025},
  doi = {10.14569/ijacsa.2025.0160341},
}
Figure