A Multi-Mechanism Seagull Optimization Algorithm Incorporating Generalized Opposition-Based Nonlinear Boundary Processing

The seagull optimization algorithm (SOA), a well-known illustration of intelligent algorithms, has recently drawn a lot of academic interest. However, it has a variety of issues including slower convergence, poorer search accuracy, the single path for pursuing optimization, and the simple propensity to slip into local optimality. This paper suggests a multi-mechanism seagull optimization algorithm (GEN−SOA) that incorporates the generalized opposition-based, adaptive nonlinear weights, and evolutionary boundary constraints to address these demerits further. These methods are balanced and promoted the population variety and the capability to conduct global and local search. Compared with SOA, PSO, SCA, SSA, and BOA on 12 well-known test functions, the experimental results demonstrate that GEN-SOA has a higher accuracy and faster convergence than the other five algorithms, and it can find the global optimal solution beyond the local optimum. Furthermore, to verify the capability of GEN−SOA to solve practical problems, this paper applied GEN−SOA to solve two standard engineering optimization design problems including a welding optimization and a pressure vessel optimization, and the experimental results showed that it has significant advantages over SOA.