Abstract: Objectives In the framework of the Discrete-Module-Beam (DMB) method and based on the genetic algorithm, a encoding technique is proposed to specifically deal with the connector stiffness optimization problem of modular box-pontoon-type offshore floating photovoltaic (OFPV) platform. Methods The DMB method is first introduced where the connector stiffness matrix is given and recovery of hydroelastic behaviour is illustrated. The genetic algorithm is then introduced including the weighted sum method and the non-dominated sorting II method. Three encoding techniques, that is, real encoding, exp-encoding and sci-encoding, and their crossover operator and mutation operator are elaborated. The equivalent zero stiffness and equivalent infinite stiffness are proposed to further refine the solution space. Results The NSGA-II algorithm gives the Pareto front under two objectives, i.e., minimizing the maximum structural shear force and minimizing the maximum structural bending moment. It is also proved that the Pareto front could be viewed as a set of results given by the weighted sum method under different weight combinations. Conclusions The developed numerical model is able to perform multi-objective-optimization on the connector stiffness of modular box-pontoon-type OFPV.