A novel hybrid RSO with WOA for single-objective and multi-objective optimization

The hybridisation process combines the strengths of various metaheuristics while simultaneously minimising their deficiencies. This work proposes a novel hybrid algorithm that integrates rat and whale-inspired metaheuristics, with the aim of achieving superior convergence without compromising computational cost. The proposed hybridisation incorporates the hunting and fighting behaviours of the rat algorithm into the prey-encircling mechanism of the whale algorithm. This study presents two versions of the hybrid algorithm: one for single-objective optimisation and the other for solving multi-objective and many-objective problems. The single-objective version is evaluated by minimising 33 benchmark single-objective functions. The multi-objective version was evaluated using the Inverted Generational Distance (IGD) metric on 10 multi-objective and 15 many-objective problems. The two proposed hybrids were applied to multivariate system identification, where the quality of the obtained models is evaluated using the Mean Squared Error (MSE) and the Coefficient of Determination (R2). The performance of the proposed single and multi-objective hybrid versions was compared with several well-established swarm-based metaheuristics and hybrid algorithms from the literature. The results illustrate the benefits of hybridising whale and rat metaheuristics in terms of convergence and computational cost.