Path planning of multiple UAVs with online changing tasks by an ORPFOA algorithm

• Published in: Engineering applications of artificial intelligence Vol. 94; p. 103807
• Main Authors: Li, Kun, Ge, Fawei, Han, Ying, Wang, Yi’an, Xu, Wensu
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.09.2020
• Subjects: Fruit fly optimization algorithm; Oilfield inspection; Path planning; Swarm intelligence algorithm; Task assignment; UAVs; Oilfield inspection; Swarm intelligence algorithm; Path planning; Fruit fly optimization algorithm; UAVs; Task assignment
• ISSN: 0952-1976; 1873-6769
• DOI: 10.1016/j.engappai.2020.103807

The unmanned aerial vehicle (UAV) is a new type oilfield inspection tool which is characterized by high flexibility, low cost and high efficiency. In the UAV based oilfield inspection technology, the path planning is an indispensable element which finds an optimal flight path for UAV to finish the inspection jobs successfully. In comparison with the other researches, our study focuses on two challenging issues: path planning of multiple UAVs by traversing a certain amount of task points in the three-dimensional environment within the required completion time, and optimizing solving for the best flight path with online changing tasks. In the research, a novel task assignment method including the initial task assignment and the task assignment with changing tasks is proposed to determine the initial task sequences of each UAV and rapidly replan task sequences after tasks change. An improved fruit fly optimization algorithm (named ORPFOA) is proposed to solve the path planning problem in both initial task sequences and new task sequences after tasks change, in which the optimal reference point and a distance cost matrix are used to reach both faster solving and higher optimizing precision for the optimal flight path. In ORPFOA, two cost functions are defined to evaluate the optimizing results in the initial phase and the new phase after task changes, respectively. A simulation model of the three-dimensional oilfield environment is established to verify the effectiveness of the proposed method in comparison with other six algorithms. [Display omitted] •A new method is proposed to generate an optimal initial path for multiple UAVs.•The task assignment with task priority and changing tasks is considered.•An improved FOA algorithm based on the optimal reference point (ORPFOA) is proposed.•A new evaluation mechanism using the cost function is proposed.