Hybrid machine learning modeling and bioinspired metaheuristic optimization of tensegrity trusses for structural design

Abstract Typical optimization approaches fail to respond to changing load conditions and structural response nonlinearity limit tensile truss design. Adaptive tensile learning and swarm-driven material redistribution improve structural adaption and efficiency in the bioinspired metaheuristic framework. The model increases convergence, minimizes material, and maintains equilibrium fidelity across load circumstances. Comparisons reveal 35% less material and 30% faster convergence than benchmark metaheuristics. These findings enable intelligent, self-regulating tensegrity systems for deployable architecture, aeronautical constructions, and adaptive robotics. The integrated framework saves about 35% of materials, improves accommodation by around 40%, and shortens convergence time by nearly 28% when compared with traditional meta-heuristic methods. It also allows quick adaptations in geometry, self-healing from damage, and dynamic balance of this stress in tension or compression. This paper opens a novel path for the next generation of intelligent bioinspired tensegrity systems and deep structural awareness for applications in deployable architecture, aerospace structures, and soft robotics.