Optimization of material composition of nonhomogeneous hollow sphere for thermal stress relaxation making use of neural network

Optimizing the composition profile of a functionally graded interlayer using a direct transcription method

“…Profiles of this form have been extensively used in simulation and optimization studies (reducing the optimization problem to a univariate optimization problem) [4,11,[26][27][28][29][30][31][32].…”

“…Setting x = {u, σ r , T } and m = {λ, G, K , α}, we formally recast the governing equations (9)(10)(11) in the form of the vector ordinary differential equation…”

“…finding the optimal phase volume fraction of each constituent of the component, all else being assumed fixed). The methods developed for this purpose range from descent techniques as applied to single-and multi-objective programming [6][7][8][9] to heuristic techniques such as those based on genetic algorithms and neural networks [10][11][12]. This paper also contributes to composition optimization by implementing techniques of the former kind.…”

Determination of Elastic Constants of Anisotropic Laminated Plates Using Elastic Waves and a Progressive Neural Network

“…The activation function used in this work is a sigmoid function. In equation (11), is the additional jump factor for overcoming the stagnation of the weight matrix, the value should be between 0 and 0)15 and can be varied during the training process. The purpose of adding this small positive value to f (zI G ) is to maintain a non-zero I G value and prevent the weight matrix from stagnation.…”

16.1 Functionally Graded Materials

“…Zimmerman and Lutz [14], Jabbari et al [15,16], Liew et al [17], Shao et al [18,19], and Hosseini et al [20] have analyzed thermoelastic problems for FG cylinders or cylindrical panels. Ootao et al [21,22], Afsar and Sekine [23], and Goupee and Vel [24] have optimized material compositions for inhomogeneous hollow circular cylinders.…”