Enhanced Adaptive Simulated Based Artificial Gorilla Troop Optimizer for Global Optimisation

Abstract

The enhancement of metaheuristic algorithms has been considered an important step in improving the solution quality of systems. In this paper, a modification to the traditional Gorilla Troop optimizer is proposed. The modification leverages the powerful properties of circle chaotic mapping and step adaptive simulation. The conventional algorithm, while effective in certain scenarios, exhibits limitations in handling complex and dynamically changing data sets. To address these shortcomings, a three-fold approach is proposed to enhance its performance. A circle chaotic mapping is integrated into the algorithm's initialization phase to enhance its sensitivity to initial conditions. The chaotic mapping effectively diversifies the search space, facilitating improved exploration and convergence to optimal solutions. Secondly, a step adaptive simulation is introduced as a means to dynamically adjust the simulation steps during runtime. Finally, the concept of adaptive simulation based on the state of the silverback gorilla (best solution) in the troop is used to simulate the exploitation phase to help the ASGTO overcome local optima entrapment and produce better solutions. The performance of the proposed ASGTO was assessed on twenty-two benchmark optimization functions and compared with the standard GTO, grey wolf optimizer (GWO), and whale optimization algorithm (WOA). The results showed that the proposed ASGTO outperformed the standard GTO, GWO, and WOA. ASGTO, GTO, GWO, and WOA attained global optimum values for 82%, 77%, 55%, and 59% of the 22 benchmark functions respectively. Consequently, the modified algorithm exhibits robustness and adaptability, making it applicable across various domains. The ASGTO is therefore recommended for adoption in solving optimization problems.