Abstract: Objective To realize the matching relationships between the dynamic requirement and functional models of maritime tasks in the manner of unmanned clusters collaboration and then accomplish the maritime operation tasks, it helps to greatly improve the operation efficiency and reduce the operator’s intervention rate. Methods Aiming at the problem of difficulties during the matching of requirement and functional models caused by dynamic properties of the requirement model of maritime unmanned cluster collaborative operation, based on the dynamical operation requirement data and logical model of unmanned equipment, taking the dynamical operation requirements and functional model of maritime unmanned cluster as the input, an improved genetic algorithm based on elite retention hierarchical selection strategy is proposed to generate the matching solution within 18 seconds. Meanwhile, this method could also dynamically generate the optimal functional model matching solution with the best performance. It helps to improve the operation efficiency of maritime unmanned cluster collaboration. Results The results show that the proposed method has the ability of dynamically generating the optimal functional matching solution considering the dynamical operation requirement of maritime unmanned cluster collaborative operation. In addition, the generated matching solution of the proposed method is better than artificial bees colony algorithm and bees algorithm. Conclusion The proposed theory and method could be applied to the research field of optimization of maritime equipment design and operation. It could improve the intelligence level of maritime equipment.