.png){kind=icon}
.png){kind=icon}
Return
Probabilistic confidence-based estimation method for structural resistance design/characteristic value
.png){kind=icon}
PRE
AI
DOI：10.1016/j.istruc.2025.109639.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The engineering structural design process depends on accurately estimating resistance design and characteristic values. However, obtaining precise estimates is challenging with limited sample sizes, potentially leading to designs that are either excessively conservative, which raise costs, or overestimate, which compromise structural safety and reliability. To address this issue, small-sample estimation methods have garnered significant attention. However, these methods often struggle to effectively regulate assurance rates and confidence levels. In response, this study introduces a probabilistic confidence-based estimation approach for forecasting structural resistance design and characteristic values. The proposed method linearizes the nonlinear relationship between resistance values and influencing factors using the least squares technique, enabling the separate estimation of two resistance components—those with known and unknown probabilistic properties—while accounting for confidence levels to ensure reliable assessments. The overall resistance design/characteristic value is then derived by integrating these two component estimates. Simulations and case studies validate the effectiveness of the method, demonstrating enhanced accuracy and improved probabilistic behavior in estimating design and characteristic values compared to traditional code methods. Additionally, the study provides charts to determine the required sample size for accurate estimations. This research advances the structural design and evaluation framework by utilizing design and characteristic values as random variables.