Abstract
Pile foundations are used to transmit construction loads deep into the ground, thus ensuring the stability of the structure. In this context, there is no question that the calculation of pile bearing capacity is important in the design of economical and safe geotechnical structures. Typically, the axial bearing capacity of piles can be determined using five approaches. At the same time, under real conditions, the effectiveness of these approaches is limited, firstly, by the restrictions inherent in their computational apparatus and, secondly, by the action of a variety of uncertain and random factors. Meanwhile, in geotechnics the uncertainties are largely unknown or very difficult to measure. In view of the above, the aim of the article is to develop a methodology for analyzing the reliability of a pile foundation taking into account uncertainties and partial factors. The methodological basis of the research is a comprehensive approach, which includes the analysis and synthesis of literary data on the topic of work, analytical and experimental research. In the process of research an algorithm based on the Monte-Carlo method and the evolutionary neural network, which allows in the process of calculating the reliability of piles to take into account such uncertainties as: physical, statistical and modeling uncertainties. The conducted tests showed that the model has a high prediction accuracy. The theoretical value of the obtained results consists in the development of the pile foundation reliability evaluation apparatus due to the use of neural network modeling. In practice, the recommendations formulated in the article can be used as a basis for conducting experiments with pile foundations in various soil sciences conditions.