Probabilistic–statistical method for assessing the technical condition of rubber bearings in building seismic isolation systems

Introduction. Seismic isolation systems based on rubber–metal bearings represent an effective engineering solution for reducing seismic loads and, consequently, structural damages to buildings during earthquakes. However, during long-term operation the materials of the bearings are subject to aging and degradation of their physical and mechanical properties, which affects the dynamic characteristics of buildings and the efficiency of seismic isolation. The absence of standardized methods for the prompt quantitative assessment of the technical condition of rubber–metal bearings under operating conditions necessitates the development of new diagnostic approaches. Aim. The study is to develop and formalize a probabilistic–statistical method for the quantitative assessment of the technical condition of rubber bearings in seismic isolation systems of buildings in operation based on the analysis of dynamic monitoring data. Materials and methods. The probabilistic–statistical method proposed by the author for assessing the technical condition of rubber–metal bearings in building seismic isolation systems is based on the analysis of dynamic monitoring data and relies on a Bayesian dynamic linear model. Within this framework, the technical condition of the seismic isolation system is described by a vector of latent state parameters comprising one or several components. The evolution of the state parameters is modeled as a stochastic process, while the relationship between the state parameters and the observed dynamic characteristics is represented using a surrogate model based on a Gaussian process. Bayesian inference is implemented using Markov chain Monte Carlo methods. Results. The paper presents a formalized probabilistic–statistical method for assessing the technical condition of rubber–metal bearings in seismic isolation systems according to categories defined in the parameter space, taking into account the normatively permissible changes in their properties and the requirements for ensuring the seismic resistance of the building. The possibility of extending the method by incorporating additional diagnostic indicators is also demonstrated. Conclusion. The proposed probabilistic–statistical method enables the transition from a qualitative interpretation of changes in a building’s dynamic response to a quantitative assessment of the technical condition of the seismic isolation system. The method can be integrated into building dynamic monitoring systems and used both for assessing the technical condition of rubber–metal bearings and for predicting their residual service life.