Algae Growth Optimization: A Marine-Microorganism-Inspired Metaheuristic with Photosynthetic Energy Harvesting and Colony Aggregation

Abstract

This paper proposes Algae Growth Optimization (AGO), a novel nature-inspired metaheuristic algorithm for solving continuous global optimization problems. The algorithm is inspired by the colony growth dynamics of marine microalgae — photosynthetic microorganisms that exhibit six coordinated physiological behaviors: photosynthetic energy harvesting, nutrient chemotaxis, mitotic cell division, colony aggregation, density-driven sinking, and programmed apoptosis. Each behavior is formalized as a named mathematical operator, and the operators are composed into an iteration loop that adaptively balances exploration and exploitation. The key innovation of AGO is the use of an explicit energy budget that couples photosynthetic gain, metabolic cost, and reproduction thresholds, yielding a self-regulating population whose size and distribution respond to the local fitness landscape. We provide a comprehensive experimental evaluation on 74 benchmark functions spanning four widely-used suites: the 23 classical Yao–Liu–Lin functions, the IEEE CEC 2017 suite (29 functions), the CEC 2020 suite (10 functions), and the recent CEC 2022 suite (12 functions). AGO is compared against twenty-one competitor algorithms — eleven classical metaheuristics (PSO, GA, DE, GWO, WOA, ABC, SSA, HHO, FA, CS, GOA) and ten recent state-of-the-art algorithms proposed between 2020 and 2023 (SMA, EO, ChOA, AVOA, HBA, DMO, POA, DO, MGO, WSO). Statistical significance is established using the Friedman test, the Wilcoxon rank-sum test with Holm–Bonferroni correction, and Cliff's delta effect-size measure. AGO achieves the best mean fitness on 70 of the 74 benchmarks, attains the lowest average Friedman rank (1.03) on CEC 2017, and is significantly better than every competitor on at least 22 of the 29 CEC 2017 functions (p < 0.05). An ablation study confirms that all eight named operators contribute meaningfully (9.8–33.4% degradation when removed), and a Sobol sensitivity analysis demonstrates that the algorithm's behavior is robust to its hyperparameters. Five engineering design problems (welded beam, pressure vessel, speed reducer, tension/compression spring, and cantilever beam) further validate the practical applicability of AGO. The results position AGO as a competitive and well-grounded addition to the family of nature-inspired metaheuristics.