Extended Fusion Algorithm of A-star and Artificial Potential Field for Three-Dimensional Path Planning

최영훈 교수의 연구실에서 인공지능 드론의 3차원 비행 경로 플래닝 기술을 연구하여 발표한 논문 "Extended Fusion Algorithm of A-star and Artificial Potential Field for Three-Dimensional Path Planning"이 국제 학술 대회인 "Joint conference of APCATS, AJSAE & AAME" 2023년 10월 호에 게재되었다.  

 

논문 사이트로 이동 

 

 

[Figure 1] Simulation of fusion algorithm

  

 

Urban Air Mobility (UAM) has been recently getting attention due to environmental and economic reasons. This has led to research on three-dimensional path planning requiring more degrees of freedom than traditional two-dimensional path planning. Traditional path planning algorithm such as A-star and artificial potential field has focused on how to avoid collisions and how to guarantee path’s optimality. A-star algorithm is, however, hard to handle unknown obstacles, and the artificial potential field could not reach to the given destination due to local minima problem. To relieve these issues, the fusion algorithm of A-star and artificial potential field was developed. 

 

 

 

[Figure 1] Comparison of algorithms with known and unknown (purple) obstacles

 

 

The algorithm is, however, mainly relevant to two-dimensional path planning problems. To deal with three-dimensional path planning cases for UAM, this paper introduces an extended fusion algorithm of A-star and artificial potential field algorithms to handle three-dimensional path planning problems. To demonstrate the proposed algorithm, numerical simulations are conducted with traditional A-star and artificial potential field algorithm.