Abstract
In this paper, a hybridization model based on culture algorithm and Artificial Bee Colony is proposed. The objective of the hybrid model is mainly to get benefit of the previous knowledge gained by predecessor foragers which help bees searching for food sources in potential positions. The proposed CB-ABC focuses on the kind of information in the belief space that should be stored to reduce promising solutions' area. Moreover, CB-ABC divides the population into two groups of individuals, one group updates by the heritage of best previous solutions and the other group updates by self-adaptive information. The performance of the new algorithm has been validated on a variety of numerical testbench functions (ranging from CEC 2005 and CEC 2017) and compared to standard ABC and other variants named ABCM, HPA and CABCA as well.
The proposed algorithm proves its success as well when applied on Wireless Sensor Network (WSN) to locate them accurately. To validate the performance of CB-ABC it is compared with Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Cuckoo Search (CS), Firefly Algorithm (FA) and Parallel Firefly Algorithm (PFA). CB-ABC shows the least average error value compared to the standard ABC as well as the other algorithms previously mentioned in the comparison. Moreover, CB-ABC succeeds in reducing the number of iterations as well as function evaluations to 17% of those of the standard ABC to 20% of those obtained by ABCM during solving WSN localization problem. CB-ABC's parametric study (heritage size and offspring ratio) has been carried out as well. (C) 2019 Elsevier B.V. All rights reserved.