Targeting the complex geometric shapes and multi-level characteristics of metamaterials, a simulation and topology optimization method based on fixed grid techniques is proposed. Users do not need to manually partition finite element meshes which conform to geometric shapes, thereby significantly reducing preprocessing time. By employing quadtree/octree based local adaptive refinement, the accuracy of structural analysis can be ensured without increasing finite element analysis computational costs. And higher resolution topology optimization results can be generated. Based on the proposed method, a multifunctional metamaterial topology optimization model considering both load bearing and heat conduction was established. Numerical implementation and verification were conducted using a triply periodic minimal surface (TPMS) composite rib-reinforced structure as the optimization object. The proposed method has been implemented in OptFuture, a domestically developed and fully autonomous CAX industrial software. In addition, to address the multilevel weight reduction requirements of metamaterial structures, OptFuture is used to achieve multi-level lattice design and filling of metamaterials, thereby further expanding the potential for lightweight design of metamaterials.