Objective: With the rising global energy demand and the urgent pursuit of sustainable energy solutions, offshore photovoltaic (PV) systems have emerged as a highly promising option. However, existing applications of fixed offshore PV structures are restricted to water depths of merely 3-5 m. This narrow range limits the large-scale development of offshore PV power generation. This has facilitated an urgent need to expand the applicable water depth range and clarify more effective structural forms. This research aims to develop an automated simulation and optimization method. This method is particularly suitable for handling problems involving complex nonlinear relationships and multiple conflicting goals, such as those encountered in the design of offshore PV structures. By addressing the combinatorial explosion problem caused by numerous design variables in the optimization process, the study focuses on parameterized finite element analysis and optimization design to enhance the performance and expand the application scope of fixed offshore PV structures. Methods: This research aims to optimize fixed offshore PV structures. Several design variables, including water depth, the number of piles, and support structure parameters, are carefully selected. The goal is to maximize structural stiffness while minimizing costs, two objectives that often conflict with each other. The Isight platform was used to integrate finite element analysis with a multi-island genetic algorithm. This integration enables the automation of the simulation and optimization process for PV structures in water depths ranging from 4 to 16 m. First, the PV structure is parameterized using Abaqus software. In the Abaqus model, the structure is modeled with beam elements, and all connections are assumed to be rigid for simplicity and computational efficiency. The bottom of the structure is constrained as fixed, which simplifies the boundary conditions while effectively capturing the main structural response characteristics. After parameterization, the multi-island genetic algorithm, integrated within the Isight platform, is used to search for the optimal combination of design variables. This algorithm divides the population into multiple sub-populations (islands), and each sub-population evolves independently for a certain number of generations. Then, individuals are exchanged between islands, which helps avoid local optima and explore a wider design space. Results: The optimized design can significantly increase the applicable water depth to 16 m while fully satisfying all design constraints. This achievement provides substantial technical support for the development of shallow-sea photovoltaics and broadens the scope of offshore PV applications. A surrogate model has been constructed based on the proposed method. Using this surrogate model greatly enhanced the efficiency of the optimization process. This eliminates the need for repetitive finite element modeling, which is time-consuming. The sum of squared residuals of the surrogate model is less than 0.1, and the root-mean-square error ranges from 0.75 to 0.9. These values indicate that the model fit can effectively capture the primary response tendencies of the structure. In terms of structural performance, the optimization process could reduce the maximum stress that the structure endures by about 45% while keeping the structural mass unchanged. This outcome not only improves the safety factor of the structure but also shows great potential for material efficiency and cost reduction in offshore PV structure design. Conclusions: This study presents an automated simulation and optimization methodology that significantly enhances the efficiency and effectiveness of offshore PV structure design. The approach is applicable to offshore PV structures, and it provides a theoretical reference for the preliminary design of other novel offshore structures, contributing to the advancement of offshore renewable energy technologies.