An Adaptive Whale Optimization Algorithm Using Gaussian Distribution Strategies and Its Application in Heterogeneous UCAVs Task Allocation

Abstract

To overcome the defect of whale optimization algorithm (WOA) being easily fallen into local optimum caused by the ill-distribution of solutions, this paper explores an adaptive WOA variant using Gaussian distribution strategies (GDSs), named GDS-WOA. In GDS-WOA, by means of one GDS, named the Gaussian estimation of distribution method, the superior population information is used to evolve the distribution scope and modify the evolution direction. Moreover, an adaptive framework is adopted to integrate the Gaussian estimation of distribution method and WOA together, in which each individual can update its position using Gaussian estimation of distribution method or WOA according to an adaptive probability parameter. When the algorithm falls into stagnation, another GDS, named Gaussian random walk, is activated to enrich the population diversity and help the algorithm get rid of the local optimum. Additionally, the greedy strategy is carried out to select the offspring from the parents and the generated candidates to fully retain the promising solutions. The GDS-WOA is benchmarked on CEC 2014 test suite, and the performance of GDS-WOA is evaluated by comparing with WOA and its promising variant IWOA, as well as other five state-of-the-art evolutionary algorithms, i.e., COA, VCS, CoBiDE, HFPSO and GWO. The statistical results demonstrate that GDS-WOA outperforms other competitors in terms of convergence efficiency and accuracy. Finally, GDS-WOA is applied to solve the optimal task allocation problem of heterogeneous unmanned combat aerial vehicles (UCAVs). To address this constrained real-world optimizing problem efficiently, the mathematical model of heterogeneous UCAVs task allocation is described with the operational effectiveness value as the objective. The validity and practicauility of the model as well as the performance of GDS-WOA for solving constrained optimization problem are demonstrated by the experimental results.

A Novel Hybrid PSO-K-Means Clustering Algorithm Using Gaussian Estimation of Distribution Method and Lévy Flight

“…The estimation of distribution algorithm (EDA) can estimate the evolution direction of the promising population using probabilistic model learning and sampling. Some studies have shown that EDA has promising performance when dealing with complex optimization problems [39] [40]. GEDM is the core component of EDA.…”

Tuna Swarm Optimization: A Novel Swarm‐Based Metaheuristic Algorithm for Global Optimization

“…The metaheuristic algorithm has the advantages of not relying on the problem model, not requiring gradient information, having strong search capability and wide applicability, and can achieve a good balance between solution quality and computational cost [ 3 ]. Therefore, the metaheuristic algorithms have been proposed to solve real-world optimization problems, such as image segmentation [ 4 , 5 ], feature selection [ 6 , 7 ], mission planning [ 8 , 9 ], parameter optimization [ 10 , 11 ], job shop scheduling [ 12 , 13 ], etc.…”

A Modified Slime Mould Algorithm for Global Optimization

“…us they are widely used for solving optimization problems, such as mission planning [4][5][6][7], image segmentation [8][9][10], feature selection [11][12][13], and parameter optimization [14][15][16][17][18]. Metaheuristic algorithms find optimal solutions by modeling physical phenomena or biological activities in nature.…”