Particle Swarm Optimization Algorithm

In this article, we introduce a nature-selection-based particle swarm optimization algorithm that operates through customizable parameters. For instance, we can use global popsize to define population size; global pop to specify the population; global c1 and global c2 to set individual best guidance coefficient and global best guidance coefficient respectively. Additionally, we can utilize global gbest_x, global gbest_y, global best_fitness, and global best_in_history to specify the x-coordinate, y-coordinate of the global best solution, optimal solution value, and optimization trajectory respectively. The algorithm implementation uses global x_min and global x_max to define lower and upper bounds for x, while global y_min and global y_max set boundaries for y. The iteration control parameters include global gen for maximum iteration count, global exetime for current iteration number, and global max_velocity for velocity limitation. Through these customizable parameters, the algorithm efficiently implements velocity updates using v = w*v + c1*rand()*(pbest-x) + c2*rand()*(gbest-x) and position updates via x = x + v, ensuring optimal convergence through nature-inspired selection mechanisms.