1. Improved A* algorithm based on a dynamic parameter and the Bezier curve.
- Author
-
Xu, Yang
- Subjects
- *
EUCLIDEAN metric , *ROBOT motion , *ALGORITHMS , *EUCLIDEAN algorithm , *EUCLIDEAN distance , *CURVES , *PARAMETERIZATION - Abstract
Within the domain of robotics and related fields, path planning algorithms have perennially posed a profoundly challenging problem, marked by an ongoing quest for refinement without attaining a universally perfect solution. Among the spectrum of path planning algorithms, A* algorithm stands out as a relatively stable and efficient approach. This paper presents improvements to the conventional A* algorithm by introducing several key modifications. Firstly, it adopts the Euclidean distance metric instead of the Manhattan distance, thereby mitigating the proclivity of the algorithm for diagonal movements. Secondly, the dynamic weight parameterization is introduced, markedly amplifying computational efficiency. In exemplar map scenarios, this augmentation yields a reduction of approximately 50% in program execution time compared to the conventional A* algorithm. Lastly, the integration of Bezier curves serves to optimize path inflections, imparting a smoother trajectory. This approach substitutes discrete grid-based path planning inflection points with continuous curves, obviating the need for abrupt halts and turns in the trajectory of the robot during actual execution. Incorporating the stability and optimality inherited from the traditional A* algorithm, this algorithm enhances program execution speed while reducing inflection points in the route, substituting them with smooth curves. The results identify an improved A* method as an important improvement in path planning, which greatly improves the computational and operational efficiency related to robot motion in production practice. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF