ENHANCED PRAIRIE DOG METAHEURISTIC OPTIMIZATION ALGORITHM FOR ENGINEERING OPTIMIZATION PROBLEMS

This paper presents an Enhanced Prairie Dog Optimization (IPDO) algorithm for solving complex engineering optimization problems. The proposed improvement integrates Lévy flight dynamics into the original PDO framework to enhance exploration-exploitation balance and accelerate convergence. The performance of IPDO is evaluated against seven established metaheuristics across four challenging civil engineering applications: (۱) discrete sizing optimization of a ۱۲۰-bar truss, (۲) structural reliability analysis of a cantilever tube, (۳) cost optimization of reinforced concrete beams, and (۴) hyperparameter tuning of a Support Vector Machine (SVM) for shear strength prediction of steel fiber-reinforced concrete. Experimental results demonstrate that IPDO consistently achieves superior solution quality, robustness, and convergence speed. Notably, in SVM hyperparameter optimization, IPDO attained the lowest mean squared error (۱.۴۸۸۱) with zero variance across runs, outperforming all competitors. The algorithm also proved highly effective in structural design and reliability problems, offering a reliable and efficient tool for real-world engineering optimization.

کلیدواژه ها:

Engineering optimization ، Improved Prairie Dog Optimization ، Metaheuristic algorithm ، Truss structures