A novel version of Jumping Spider Optimization for Global Optimization and Real-World Engineering Issues

Drawing inspiration from the hunting prowess of jumping spiders, the jumping spider optimization algorithm (JSOA) was developed as an intelligent optimization technique known for its commendable convergence speed. Despite its efficiency, JSOA is susceptible to getting ensnared in local optima. To overcome this limitation, this study proposes a novel variant named the multi-strategy improved jumping spider optimization algorithm (MSIJSOA). MSIJSOA simulates an additional hunting behavior observed in jumping spiders, which involves scent-based prey detection followed by pursuit. To emulate this, the algorithm incorporates parameters defining scent ranges, delineating the predatory process into exploration and exploitation stages. Additionally, it integrates differential pheromone fusion and natural selection strategies to facilitate escape from local optima. To validate the effectiveness and robustness of MSIJSOA, comprehensive evaluations are conducted using CEC2017 and CEC2020 benchmark sets, benchmarking against both classical and state-of-the-art optimization algorithms. Results from experimental trials showcase MSIJSOA's superior convergence speed and solution quality. Furthermore, the algorithm demonstrates its efficacy by successfully tackling eight real-world engineering problems. The experiments also highlight MSIJSOA's robustness and versatility in addressing a wide array of complex challenges.