Multi-Objective Optimization of Rein-Forced Concrete Buildings using NSGA-II

With many design elements, goals, and limitations, risk-based designs are intricate and frequently computationally demanding. Single-objective optimisation is frequently used in structural fire engineering for risk-based designs; nevertheless, this method may converge to a local optimum and produce unfeasible design solutions. These problems can be resolved for risk-based designs by putting multi-objective optimisation (MOO) methods into practice. A technique for optimising reinforced concrete (RC) mix proportions with regard to various reliability and economic goals—or, conversely, failure probability—is presented in this research. This method forecasts the cost of concrete using a method based on linear regression and the compressive strength of concrete over a 25-day curing period using a quadratic generalised ridge regression model. With NSGA II, dependable pof with non-dominated solutions are produced for a range of compressive strength requirements. Analysis is done on Pareto-optimal fronts that have evolved by changing the constraints for compressive strength and failure probability. It is found that, up to a certain degree, the cost increases nominally as the chance of failure decreases for a particular strength in compression need. Below that threshold, however, the price of concrete increases significantly.