Estimating the deformation of micropile stabilized footings by GEP approach

Micropiles possess a remarkable adaptability that allows them to function effectively as both a novel foundation and a sustainable solution for existing foundations, offering numerous advantages such as exceptional load-carrying capacity and minimal disturbance to the underlying subgrade. They have the potential to revive essential infrastructures, historic structures, or distressed structures. The primary objective of this investigation is to explore the in-situ reinforcement method, specifically vertical micropile reinforcement, for enhancing foundation performance. The research method employed involves a sequence of laboratory model tests on a square footing surrounded by micropiles, and subjected to vertical concentric loading. Additionally, the Gene expression programming (GEP) technique was employed to model the micropile deformation, and by conducting numerous trials, the optimal trial yielded an R ² of 0.9595 and a mean absolute error of 1.1135. The GEP hyperparameter tuning was achieved using trial and access, which indicated that the optimal head size was 10, the number of genes was 3, and the number of chromosomes was 200. The results indicate that 90% and 75% of the predicted values fall within the 20% relative error range for the training and validation data, respectively. Furthermore, a parametric analysis was conducted based on various geometrical and soil parameters to determine their impact. The results demonstrate that the placement of micropiles around a footing can significantly enhance its settlement characteristics and load-bearing capacity.