Artificial neural network methodology for the estimation of ground enhancing compounds resistance

The work presented in this study aims to develop a methodological approach for estimating the ground resistance of several grounding systems, embedded in various ground enhancing compounds, using artificial neural networks (ANNs). The ANN training is based on field measurements that have been performed in Greece during last years. The methodology uses as input variables measurements of soil resistivity within various depths and of rainfall height during some periods of time, like last week and last month and estimates the ground resistance value of the tested rods, based on an ANN. This work comprises two scenarios in which, several ANN training algorithms are applied and an optimisation process is performed regarding the values of parameters, such as the number of neurons, the activation functions combination and so on. Each training algorithm is compared to the others, based on the coefficient of determination between the experimental and estimated values for the test set and the algorithm with the best results is highlighted for the estimation of ground resistance value, formed by the ground enhancing compounds under various weather conditions.