A parallel cooperative coevolutionary multi-verse algorithm for large-scale multi-objective UAV path planning problems

Path planning of Unmanned Aerial Vehicles (UAVs) in complex environments with high dimensionality attempts to search for waypoint sequences always of increased size. Such a challenging task can be considered as a multi-objective Large-Scale Global Optimization (LSGO) problem where efficient solving...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Memetic computing Jg. 17; H. 4; S. 44
Hauptverfasser: Jarray, Raja, Bouallègue, Soufiene
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2025
Springer Nature B.V
Schlagworte:
Algorithms
Applications of Mathematics
Artificial Intelligence
Bioinformatics
Collaboration
Collision avoidance
Complex Systems
Complexity
Computing time
Control
Decision making
Decomposition
Energy consumption
Engineering
Feature selection
Genetic algorithms
Global optimization
Mathematical and Computational Engineering
Mechatronics
Message passing
Multiple objective analysis
Optimization algorithms
Path planning
Planning
Regular Research Paper
Robotics
Unmanned aerial vehicles
Variance analysis
Parallel computing
Multi-core CPU allocation
Unmanned aerial vehicles
Path planning
Large-scale multi-objective optimization
Cooperative coevolutionary decomposition
ISSN:1865-9284, 1865-9292
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Path planning of Unmanned Aerial Vehicles (UAVs) in complex environments with high dimensionality attempts to search for waypoint sequences always of increased size. Such a challenging task can be considered as a multi-objective Large-Scale Global Optimization (LSGO) problem where efficient solving requires more sophisticated algorithms. In this paper, a novel Parallel Cooperative Coevolutionary Multi-Objective Multi-Verse Optimization (PCCMOMVO) algorithm is developed and successfully applied. In this Cooperative Coevolutionary (CC) framework, a MOMVO algorithm is considered to design an improved subcomponent optimizer based on an allocated multi-core CPU architecture and a Message Passing Interface (MPI). The MOMVO population is divided into sub-populations, called species, where each of them is responsible for optimizing a subcomponent of the LSGO problem according to the “divide-and-conquer” concept of CC framework. To form a complete candidate solution of the whole LSGO path planning problem, each species shares a number of representative solutions selected from the Pareto non-dominated ones found so far. The selection process is based on the Pareto ranks achieved by the use of a Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). A parallelization mechanism based on a simple yet efficient master–slave model and allocated MPI is implemented to provide acceleration in computation runtime. Demonstrative results and ANOVA tests are presented over planning scenarios with increased complexity to demonstrate the superiority and effectiveness of the PCCMOMVO algorithm. Extensive evaluations and comparisons are carried out in terms of collision-avoidance capabilities, path shortness and efficiency against the curse of dimensionality and computational time consumption.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1865-9284
1865-9292
DOI:10.1007/s12293-025-00473-3