Slope stability analysis of heavy-haul freight corridor using novel machine learning approach

In response to the increasing demand for freight transportation, specialized heavy-haul railway lines were constructed to accommodate larger trains carrying heavier loads. This led to higher productivity and lower unit prices. However, heavy-haul routes often require significant investment, necessitating careful risk assessment during the design phase. This study introduces a novel machine learning approach for slope stability analysis of a heavy haul freight corridor. The presents approach employs advanced computational techniques, such as artificial neural networks (ANN), bayesian neural networks (BNN), convolutional neural networks (CNN), and deep neural networks (DNN), which are used to evaluate the slope stability of a heavy haul freight corridor (HHRC) in terms of serviceability. The study focused on a 12.293 m high embankment on the Indian Railways (IR) HHRC. During training and testing, the proposed computational models were compared using various evaluation metrics (i.e., TIC, sMAPE, R2, RRSE, RAE, IA, and LMI). In the testing phase, the developed ANN outperformed the other model and achieved an accuracy of TIC = 0.0076 and LMI = 0.9872. Therefore, monotonicity analysis was performed using the best-performing model in the subsequent step.