Modified Particle Swarm Optimization for Engineering Optimization Problems and UAV Path Planning

Optimization problems aim to identify the best solution from a wide range of possibilities. The Particle Swarm Optimization (PSO) algorithm is widely recognized for its simplicity and efficiency, but it suffers from issues such as local optima trapping and degraded performance in high-dimensional problems. To address these limitations, this paper proposes a modified PSO algorithm (MPSO). The MPSO incorporates several novel strategies: a Sigmoid-based nonlinear inertia weight decay function, which supports global exploration in the early stages and local refinement in later stages; a non-uniform mutation operator, which amplifies perturbation to enhance global search in the early phases and reduces perturbation to guide convergence in the later phases; and a sine-cosine disturbance strategy, which boosts solution diversity and accelerates global optimization and convergence. Compared to eight PSO variants, MPSO demonstrates superior performance across various benchmark functions in the CEC2017 and CEC2022 suites, particularly excelling in high-dimensional problems. Most of the test functions show better results than the competing algorithms. Finally, MPSO is applied to five engineering optimization problems and UAV path planning tasks, where the experimental results confirm its effectiveness in real-world applications. The algorithm consistently finds high-quality solutions, highlighting its exceptional performance and practical value in solving complex optimization challenges.