Paper Info
Title
Robot path planning based on genetic algorithm with hybrid initialization method
Abstract
The research of path planning method based on genetic algorithm (GA) for the mobile robot has received much attention in recent years. GA, as one evolutionary computation model, mimics the process of natural evolution and genetics. The quality of the initial population plays an essential role in improving the performance of GA. However, when GA based on a random initialization method is applied to path planning problems, it will lead to the emergence of infeasible solutions and reduce the performance of the algorithm. A novel GA with a hybrid initialization method, termed NGA, is proposed to solve this problem in this paper. In the initial population, NGA first randomly selects three free grids as intermediate nodes. Then, a part of the population uses a random initialization method to obtain the complete path. The other part of the population obtains the complete path using a greedy-related method. Finally, according to the actual situation, the redundant nodes or duplicate paths in the path are deleted to avoid the redundant paths. In addition, the deletion operation and the reverse operation are also introduced to the NGA iteration process to prevent the algorithm from falling into the local optimum. Simulation experiments are carried out with other algorithms to verify the effectiveness of the NGA. Simulation results show that NGA is superior to other algorithms in convergence accuracy, optimization ability, and success rate. Besides, NGA can generate the optimal feasible paths in complex environments.
Year
DOI
Venue
2022
10.3233/JIFS-211423
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
Keywords
DocType
Volume
Path planning, mobile robot, genetic algorithm, initial population
Journal
42
Issue
ISSN
Citations 
3
1064-1246
0
PageRank 
References 
Authors
0.34
0
5
Name
Order
Citations
PageRank
Zhaojun Zhang131.43
Rui Lu200.34
Minglong Zhao320.73
Shengyang Luan464.85
Ming Bu500.34