An Inverse Technique to Estimate the Heat Flux of a Slab with Transient Heat Conduction

The present work concerns the application of Artificial Neural Networks (ANN) and Genetic Algorithm (GA)-based inverse technique on a transient heat transfer problem. The proposed methodology has been demonstrated on a two-dimensional heat slab to estimate linearly varying heat flux constants using space–time temperature response. Initial temperatures and thermophysical properties are assigned. Different heat flux is specified at the right and bottom surfaces of the slab. The input–output (heat flux-temperature) data set of the slab is obtained using MATLAB [1]. This is used to train the ANN network, which acts as a proxy model. The synthetic experimental temperature data set are generated from the analytical method. In the inverse problem, the GA is employed to generate the samples and minimize the objective function to estimate the constants (S1 and S2) of the heat flux function Q = S(t). The robustness of the methodology is examined for different noise levels.