Genetic Algorithms, a relatively new technique of optimization, is used for the optimum design of prestressed concrete beam. Prismatic unsymmetrical I-section girders, simply supported at the ends and subjected to a uniformly distributed load are considered for the optimum design. The total cost to be minimized includes the cost of high strength concrete, high tensile steel and a cost to reflect the cost associated with non-tensioned steel, anchorages, form, labour work, curing, testing etc. as a function of perimeter of the beam. The design problem is formulated as a non-linear mathematical programming problem to satisfy the specifications of IS:1343-1980. The effectiveness of the present method is proved by comparing it with sequential unconstrained minimization technique. Limited parametric studies have been carried out to highlight the influence of various parameters, such as cost coefficients, generation number, crossover probability and mutation probability, on the optimum cost.
|Journal of Structural Engineering (Madras)
|Published - 01-10-2004
All Science Journal Classification (ASJC) codes