A Probabilistic Minimax Estimation Framework for Structural Reliability Prediction

Authors

  • Tanveer Ahmad Wani Author
  • Dr. Pareshkumar N. Patel Author
  • Swati Sucharita Author
  • Pratibha Sharma Author
  • S N V S SS T Murty Author
  • Dr. Shanthi Vairavan Author
  • Amol Barde Author

Keywords:

Probabilistic minimax estimation, structural reliability, uncertainty quantification, variational inequality, saddle-point optimization, robust estimation, stochastic modeling.

Abstract

The prediction of structural reliability in uncertain operational conditions is still a mathematically difficult task because of the presence of stochastic loading, variation of materials, and insufficiently known probabilities. In this paper, a new approach based on the theory of stochastic modeling, minimax optimization, and variational analysis is used to provide a robust prediction of structural reliability. A minimax estimation problem is formulated using a saddle point approach where the optimal estimator finds the minimum of the maximum expected loss over the class of possible distributions. A variational inequality approach is used to derive a reliable structural reliability estimator, whereas existence, uniqueness, and convergence properties of this solution are established using the theory. A steel beam structural system is chosen for numerical demonstration using a Gaussian uncertainty model for loading and random variation of material parameters in 10,000 simulations. Improved results of the structural reliability estimator using the proposed framework are demonstrated, with a mean square error of 0.0064, worst case risk of 0.034, and reliability accuracy of 97.8%. Additionally, the suggested model has a robustness index of 95.9% with a distribution shift error of 0.0126, thus exhibiting improved performance in terms of dealing with uncertainties. According to the convergence analysis, it is stable in optimizing the problem as its objective function decreased from 0.084 to 0.011 after 50 iterations. The suggested probabilistic minimax formulation has provided a mathematically sound methodology to predict uncertainty-resilient structural reliability.

Downloads

Download data is not yet available.

Downloads

Published

2026-09-22

How to Cite

A Probabilistic Minimax Estimation Framework for Structural Reliability Prediction . (2026). Minimax Theory and Its Applications, 11(2s), 116-127. https://journalmta.com/index.php/jmta/article/view/403