Genetic algorithms-based methodology for reliability-based structural optimization

Most of the practical optimization procedures are based on fixed, component level, code safety factors and hence are almost deterministic in nature. Reliability-based structural optimization (RBSO) procedures use component and system level reliabilities as explicit design constraints and thus enable the designer to arrive at real optimum solutions by balancing / compromising the total cost and the safety level. The RBSO formulations can be made attractive to professionals by reducing the computational complexities and making the structural modeling as realistic as possible at the same time. This paper makes an attempt towards such a goal, by reconstituting the RBSO formulation using genetic algorithms. The genetic modeling and penalty-based unconstrained formulations are discussed in the case of pin jointed structures. Target reliability-based optimization strategies are proposed, considering the limit states pertaining to both service and collapse stages. Advanced First Order Second Moment (AFOSM) method is used for reliability assessment. Two examples are presented to demonstrate the robustness and efficiency of the proposed methodology in arriving at optimum solutions with the prescribed target reliability levels. The results are compared with those available in literature where mathematical optimization techniques were used for RBSO.