Enhanced Particle Swarm Optimization Algorithm for Constrained Optimization Problems

This article introduces an enhanced particle swarm optimization (PSO) algorithm specifically designed for solving constrained optimization problems. The algorithm incorporates novel strategies including dynamic penalty functions and feasibility-based selection mechanisms to improve search efficiency. Key technical implementations involve modifying the velocity update equation with constraint-aware coefficients and integrating a repair operator for infeasible solutions. Through these enhancements, the algorithm achieves better balance between exploration and exploitation while maintaining constraint satisfaction. The implementation features adaptive parameter tuning using gradient information from constraint violations, and utilizes archive-based elitism to preserve feasible solutions. This approach demonstrates significant improvements in handling complex constraint landscapes compared to standard PSO variants, providing practical value for engineering applications requiring robust optimization under constraints.