An approach to generating physically realistic data for crane safety monitoring systems based on finite element modeling

Abstract

This article focuses on the stability of monorail mobile cranes, particularly tipping and sliding stability coefficients. Traditional static models, which assume rigid structures, fail to account for deformation effects, leading to non-conservative and unreliable stability assessments. This makes them unsuitable for generating data used in AI-based safety monitoring systems. The aim of the study is to develop and substantiate an approach to generating data for assessing the stability of monorail traveling cranes based on finite element modeling, taking into account structural deformation effects. To achieve this aim, the following tasks are addressed in the paper: developing an approach to generating datasets for assessing the stability of a monorail crane based on finite element modeling; validating the proposed approach through a quantitative comparative analysis of the generated data with data obtained using the static approach; demonstrating the statistical significance of the identified differences. Methods used: experimental design, Euler-Bernoulli beam theory, statistical distribution analysis, and numerical simulation. Results show that FEM produces significantly different outcomes. The average tipping coefficient β decreased by 2.4% (from 1.398 to 1.365), and the sliding coefficient γ dropped nearly 50% (from 1.87 to 0.85), with high statistical significance (p < 0.001). Conclusions: FEM captures geometric nonlinearity, showing that elastic deformation alters the load path, increasing tipping moments and reducing stability margins. Unlike static models, the proposed method identifies unsafe operating zones previously misclassified as safe. Practical application: the model can generate synthetic data to train machine learning systems for real-time crane stability prediction, enhancing the safety of AI-driven monitoring technologies.