Mathematical modeling and optimization of intelligent systems using a hybrid PSO-GWO algorithm: A minx J(x) approach

In this paper, a comparative analysis of Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), and a hybrid PSO-GWO algorithm for the solution of complex optimization problems have been presented.The hybrid algorithm consists of the exploitation strength of GWO and the exploration capabilities of PSO while combining both together to surmount the failure of standalone algorithms like slow convergence in GWO and premature convergence in the PSO.The algorithms were evaluated in terms of convergence speed, robustness, and accuracy, on a series of benchmark functions (Sphere, Rastrigin, Ackley, Rosenbrock, and Griewank).The results of simulations indicate that the notion of a hybrid PSO-GWO algorithm is always better compared to standalone PSO as well as GWO for lower mean square errors (MSE) and quicker convergence to global optimum for different optimization landscapes.Finally, the hybrid approach took advantage of the best in all three aspects to optimize multimodal, non-convex, and deceptive functions with both reliable and robust performance that was superior to that of local minimum avoidance algorithms.This research shows that the hybrid PSO-GWO algorithm is an efficient tool for robust optimization in the real-world systems of dynamic, complex systems.Future work is to extend the algorithm's adaptability to real-world constraints, and dynamic parameter adjustment, and integrate it within domain-specific heuristics to improve the optimization in engineering and automation tasks.