Topology and Sizing Optimization of Truss Structures Using Adaptive Genetic Algorithm with Node Matrix Encoding

An adaptive genetic algorithm based on a normal sparse node matrix encoding approach is developed to handle the topology and sizing optimization of truss structure. New crossover and mutation operators are developed based on matrix chromosome. Self-adaptive strategy is utilized to enhance the robustness of genetic algorithm. Kinematically instable and structurally invalid individuals are filtered and repaired before evaluation to save computation time. Two truss examples are investigated for verification and validation. Numeric experiments show that the convergence and robustness of the algorithm developed in this paper are better than those of MIGA (multi-island genetic algorithm), especially in complicated examples.