Abstract:
In this research, the design optimization of structural concrete beams using Artificial neural networks (ANN) technique is presented .The optimization techniques in general enable designers to find the best design for the structure under consideration. The best design of structure means the most economic structure without impairing the functional purposes the structure is supposed to serve.
In this research : simple reinforced concrete beam was designed according to the requirements of the ACI 318-08 code.
The cost was calculated by a function comprises the cost of concrete, the cost of formwork and the cost of reinforcement.
A computer models have been developed for the structural design optimization of reinforced simple beams using NEURO SHELL-2 software. The results of these optimization model are compared with the results of another classical optimization model . The classical optimization models were developed based on EXCEL software spreadsheets which use the generalized reduced gradient (GRG) classical optimization method
The results obtained by using the classical optimization techniques(GRG) and the heuristic techniques (ANN) were carried out in this research.
The optimum solution for the cost of RC simple beam numerical example using the ANN model. Is 635.81 SDG and when using the spreadsheet classical model it is 633.45 SDG. This means that the minimum cost using the spreadsheet classical model is less than that for the ANN model by about 0.37 %.
This comparison showed the superiority of the ANN technique over the classical models technique with regard to our numerical example.
