Lanternfish Optimization Algorithm: A Novel Nature-inspired Metaheuristic for Efficiently Solving Constrained Engineering Optimization Problems

A novel population-based metaheuristic algorithm, Lanternfish Optimization Algorithm (LOA), inspired by the diel vertical migration and bioluminescent communication behaviours of lanternfish (Myctophidae), aimed at achieving an adaptive and robust balance between global exploration and local exploitation in complex optimization landscapes, is proposed.LOA incorporates two biologically grounded mechanisms: nighttime ascent, enabling largescale stochastic movements toward distant and promising regions guided by the global best solution, which enhances population diversity and prevents premature convergence; and daytime aggregation with photophore signalling, performing precise local search by directing individuals toward the population mean and global best, while adaptive decaying perturbations refine solutions in a cooperative manner.Results of extensive simulations conducted on 22 constrained real-world benchmark problems from the CEC 2011 suite, comparing LOA with nine state-of-the-art metaheuristics including GA, PSO, TLBO, WOA, MVO, RSA, TSA, GSA, and WSO demonstrate that LOA achieves the first rank in 17 functions, exhibits low standard deviations, compact solution distributions, and superior convergence stability, highlighting its robustness, precision, and reliability.Statistical validation using the Wilcoxon signed-rank test (p < 0.05) confirms the significance of its performance improvements.The algorithm's dual-phase design, adaptive perturbation mechanism, and population-level coordination collectively enable efficient handling of nonlinear, multimodal, high-dimensional, and constrained problems, establishing LOA as a biologically inspired, computationally effective, and highly competitive optimization framework, representing a substantial advancement.