Abstract: Fiber reinforced polymer (FRP) photovoltaic support structures exhibit broad application prospects due to their advantages of light weight, high bearing capacity, and excellent corrosion resistance. However, the design of these structures mostly relies on engineering experience and iterative trial-and-error calculations, making it challenging to attain optimal structural parameters. To address this issue, an optimal design method for FRP photovoltaic support structures was proposed. A parametric finite element model of these structures was established using ANSYS. Four strategies, namely Sobol initialization, opposition-based learning, trend-aware mechanism, and adaptive growth, were employed to enhance the optimization capability of the phototropic growth algorithm (PGA), leading to the proposal of an improved phototropic growth algorithm (IPGA). Incorporating constraints on the slenderness ratio, deformation, and stability bearing capacity of FRP members, the IPGA was used to optimize the structural shape, size, and brace layout for lightweight design. Validating in practical engineering case indicates that the parametric model accurately reflects the mechanical behavior of structures. In terms of FRP photovoltaic support optimization, the IPGA achieved a 6.05% material saving compared with the PGA, and outperformed the genetic algorithm, grey wolf algorithm, holistic swarm algorithm, and chaotic evolution algorithm in computational efficiency and optimization ability. The proposed optimal design method proved to be efficient and reliable, saving 27.73% of FRP material compared with the original scheme while ensuring structural safety.