Jian Zhou, Xiaoyou He, Taihong Lv, Hu He, and Yuhe Qiu
Coverage path planning, decomposition, input–output direction
The traditional unit decomposition method for full coverage path planning in static environments overlooked the robot’s turning weight and failed to consider the input–output direction of sub-regions when constructing the travelling salesman problem (TSP). This led to excessive redundancy and unnecessary turns in the final path. This paper presents improvements to address these issues. Firstly, it introduced an innovative decomposition method to partition the entire workspace, considering the input–output direction of each partition when modelling the partition coverage sequence. Subsequently, it utilised a simulated annealing algorithm to determine the optimal coverage sequence. To plan connections between adjacent partition paths, the A- star algorithm was employed, effectively reducing the movement path. Mathematical simulation results demonstrate that this algorithm outperforms the Contour–Boustrophedon method in overall performance. Additionally, as the number of internal obstacles increases, the enhanced algorithm exhibits more significant performance improvements. When the number of internal obstacles reaches 10, the overall cost is optimised by 26%. Finally, real flight tests using the DJI M300 and China Mobile Lingyun platform validate the effectiveness of this algorithm.
Important Links:
Go Back