An Enhanced Chicken Swarm Optimization Algorithm Using Gaussian and Tent Chaotic Map Functions

Nature-inspired optimization algorithms have proven effective in addressing complex optimization problems, but they often suffer from premature convergence to local optima. Chicken Swarm Optimization (CSO), modeled after the hierarchical behavior of chickens, is one such algorithm that, despite its strengths, can stagnate due to poor exploration dynamics. This study proposes an Enhanced Chicken Swarm Optimization (ECSO) algorithm that integrates chaotic map functions, specifically Gaussian and Tent maps, to improve its exploration capabilities and mitigate premature convergence. The developed enhancements dynamically influence the movement updates of roosters and hens, significantly improving the algorithm’s ability to discover globally optimal solutions. The ECSO is applied to optimise CNN in a forensic recognition task. Simulation results indicate that ECSO exhibits superior convergence behavior and search space coverage compared to the standard CNN and CSO optimized CNN. The developed algorithm demonstrates improved performance in both recognition accuracy and computational efficiency, validating its suitability for real-world forensic tasks.