Local Path Planning for Mobile Robots Based on Fuzzy Dynamic Window Algorithm

Ying Sun, Wenlu Wang, Manman Xu, Li Huang, Kangjing Shi, Chunlong Zou, Baojia ChenSensors 2023 · DOI

TL;DR

This paper enhances the Dynamic Window Approach (DWA) for mobile robot path planning by integrating fuzzy logic to dynamically adjust weighting parameters. This fuzzy DWA, combined with a global path guidance strategy, improves obstacle avoidance in dynamic environments, reduces planning time and path length, and mitigates the tendency to get stuck in local optima.

What Problem Does This Paper Address?

Mobile robots are increasingly used in complex, dynamic environments for tasks like navigation, inspection, and logistics. A critical component of their autonomy is path planning, which involves both global (overall route) and local (real-time obstacle avoidance) strategies. The standard Dynamic Window Approach (DWA) is effective for local planning but struggles with dynamic environments and can lead to suboptimal paths by getting trapped in local minima. This is often due to fixed weighting coefficients in its evaluation function, which don't adapt well to changing conditions. Furthermore, DWA's local focus can cause it to deviate significantly from a pre-planned global path. This paper aims to address these limitations by making DWA more adaptive and ensuring it stays closer to a global optimal path.

Methods

The core of this work is an improved Dynamic Window Approach (DWA) that incorporates two main innovations:

  • Fuzzy Logic for Dynamic Weight Adjustment: Instead of using fixed weights ($\alpha$, $\beta$, $\gamma$) in the DWA's trajectory evaluation function, this paper introduces a fuzzy controller. This controller takes the robot's distance to the goal and the nearest obstacle as inputs. Based on these inputs, it dynamically adjusts the weights $\alpha$ (target azimuth), $\beta$ (obstacle distance), and $\gamma$ (velocity) to better suit the current environmental situation. This allows the robot to prioritize different aspects of path planning (e.g., speed vs. safety) as needed.
  • Global Path Guidance with Sub-Target Points: To prevent the robot from deviating too far from an optimal global route, the paper integrates a global path planning strategy. The global path (generated by an A* algorithm) is divided into segments, and key "sub-target points" (like corners or points away from obstacles) are identified. The fuzzy DWA then aims to reach these sub-target points, effectively guiding the robot back towards the global path after local obstacle avoidance maneuvers.

The paper also details the standard DWA components, including kinematic and dynamic constraints, velocity space sampling, and the trajectory evaluation function, before explaining how these are modified.

Datasets

Not discussed in the paper. The experiments were conducted using simulations in MATLAB.

Key Insights and Ideas

  • Adaptive Planning is Crucial: Fixed parameters in path planning algorithms are a significant bottleneck in dynamic and uncertain environments. Fuzzy logic offers a flexible way to adapt these parameters in real-time.
  • Bridging Local and Global: Simply optimizing local obstacle avoidance isn't enough. Integrating local planning with a global path strategy is essential for efficient and optimal navigation over longer distances.
  • Sub-Target Points as Navigational Anchors: Using key points from a global path as intermediate goals for local planners helps maintain overall path coherence and prevents excessive deviation.
  • Fuzzy Logic for Intelligent Trade-offs: The fuzzy controller effectively balances competing objectives (e.g., speed vs. safety) by dynamically adjusting the influence of different evaluation criteria.

Key Results

The proposed fuzzy DWA algorithm demonstrated significant improvements over the standard DWA and an A* fusion DWA algorithm in simulations:

  • Reduced Planning Time: The enhanced algorithm cut planning time by 25.6% (compared to standard DWA) and 13.8% (compared to A* fusion DWA) in unidirectional motion scenarios.
  • Reduced Path Length: Path length was reduced by 10.2% (vs. standard DWA) and 3.7% (vs. A* fusion DWA) in unidirectional motion.
  • Improved Obstacle Avoidance: The algorithm successfully avoided static and dynamic obstacles in various scenarios, including complex round-trip motions.
  • Smoother Trajectories: The fuzzy DWA resulted in smoother attitude angle changes and less frequent acceleration/deceleration, leading to more efficient and less energy-consuming motion.
  • Mitigation of Local Optima: The combination of fuzzy weight adjustment and global path guidance made it difficult for the algorithm to get stuck in local optima.

Limitations and Future Work

  • Limitations:
    • The paper acknowledges that the standard DWA can struggle with "L"- or "C"-shaped obstacles and that prioritizing wider channels can increase path length. While the proposed method aims to mitigate these, the paper doesn't explicitly quantify its performance on these specific challenging shapes.
  • The simulation environment, while dynamic, might not fully capture all real-world complexities.
  • Future Work:
    • The authors suggest further research into optimizing sub-functions and dynamically adjusting weight coefficients to enhance local obstacle avoidance and handle more complex and variable dynamic environments.
  • The paper implies that the integration of fuzzy logic and global path guidance is a promising direction for future research in mobile robot path planning.

Summary updated

Abstract

Due to the increased employment of robots in modern society, path planning methods based on human-robot collaborative mobile robots have been the subject of research in both academia and industry. The dynamic window approach used in the research of the robot local path planning problem involves a mixture of fixed weight coefficients, which makes it hard to deal with the changing dynamic environment and the issue of the sub-optimal global planning paths that arise after local obstacle avoidance. By dynamically modifying the combination of weight coefficients, we propose, in this research, the use of fuzzy control logic to optimize the evaluation function's sub-functions and enhance the algorithm's performance through the safe and dynamic avoidance of obstacles. The global path is introduced to enhance the dynamic window technique's ability to plan globally, and important points on the global path are selected as key sub-target sites for the local motion planning phase of the dynamic window technique. The motion position changes after local obstacle avoidance to keep the mobile robot on the intended global path. According to the simulation results, the enhanced dynamic window algorithm cuts planning time and path length by 16% and 5%, respectively, while maintaining good obstacle avoidance and considering a better global path in the face of various dynamic environments. It is difficult to achieve a local optimum using this algorithm.
BibTeX
@inproceedings{sun2023local,
  author = {Ying Sun and Wenlu Wang and Manman Xu and Li Huang and Kangjing Shi and Chunlong Zou and Baojia Chen},
  title = {{Local Path Planning for Mobile Robots Based on Fuzzy Dynamic Window Algorithm}},
  booktitle = {Sensors},
  year = {2023},
  doi = {10.3390/s23198260},
}
Figure