A Modified Northern Goshawk Optimization Algorithm for Truss Design Problems

Northern goshawk optimization, one of the metaheuristic algorithms recently proposed for solving optimization problems, is an optimization algorithm inspired by the behavior of the northern goshawk in nature. In this study, a fitness distance balance (FDB) strategy was incorporated into the northern goshawk optimization algorithm to enhance its performance. This strategy strengthens the algorithm’s search capability and improves solution quality. Consequently, the algorithm developed with the fitness distance balance strategy has been named the modified northern goshawk optimization algorithm (MNGO). The performance of the proposed MNGO was analyzed using fifteen different benchmark functions, taking into account population size and the number of dimensions. The experimental results obtained are presented along with convergence plots. Furthermore, the proposed MNGO was compared with some other algorithms in the literature; convergence plots, box plots, and radar charts were used, and statistical analyses were also conducted. As a result of these analyses, it was observed that the performance of the MNGO improved as the number of dimensions increased. In addition, the performance of the proposed MNGO was examined statistically. Furthermore, the proposed MNGO algorithm was applied to challenging and complex 10-bar, 37-bar, 52-bar, 72-bar, and 200-bar truss problems, which are real-world engineering problems. When evaluating all 10-, 30-, 50-, and 100-dimensional problems based on average objective-function values, it was observed that MNGO achieved better results than NGO in 70% of the compared cases. When considering the ranking of the benchmark problems, the NGO algorithm ranked first for the 10-dimensional problems, while the proposed MNGO ranked first for the 30-, 50-, and 100-dimensional problems. Based on these results, it was observed that MNGO demonstrated competitive performance among the compared algorithms.