Development of an ANN controlled genetic algorithm for the numerical validation of a model for superplastic deformation

Constitutive equations that describe material behaviour under different deformation conditions involve several parameters and are often transcendental, making numerical validation against experimental data by conventional curve-fitting techniques rather difficult. Further, experimental errors in the data add to the complication in obtaining a good fit with the theoretical equation. In this context, the numerical validation of experimental data pertaining to superplastic deformation in metals and alloys, described by a grain/interphase boundary sliding controlled flow approach [1], has been considered. Earlier attempts involving conventional methods often required subjective and over-simplifying assumptions. This paper introduces the concept of "artificially intelligent genetic algorithms", in the form of an artificial neural network, to fit the data pertaining to the superplastic deformation of several metals and alloys. We propose to utilize the concept of artificial neural networks with genetic algorithms to transform a randomly generated initial set of populations of suggested solutions to a final set of populations that contains suggested solutions approximating the actual one. The fundamental concept of this paper lies in capturing the various intuitive strategies of the human brain into neural networks, which may help the genetic algorithm to evolve its population in a more lucrative manner. A carefully chosen fitness function acts in the capacity of a yardstick to appraise the quality of each "chromosome" to aid the selection phase. In conjunction with the migration phase, we employ various genetic operators and the chosen fitness function, to expedite the transformation of the initial population towards the solution. We have simulated the suggested method on a 48-node SGI Origin-2000® platform along with the utilities provided by MATLAB® software for simulating the neural networks and the results are extremely encouraging.